Astronomers peering across the universe think they’ve caught a dozen quasars—extremely bright and distant objects powered by ravenous supermassive black holes at the centers of ancient galaxies—in a disappearing act. Or at least transitioning into their quiescent and dimmer counterparts: galaxies with starving black holes at their cores. The surprising find has astronomers asking whether these objects are shutting down permanently or simply flickering out for the time being.

Last year Stephanie LaMassa from NASA Goddard Space Flight Center (then at Yale University) discovered the greatest change in luminosity ever detected in a quasar. She was digging through data from the Sloan Digital Sky Survey when she found that a quasar had dimmed in brightness by a factor of six in just 10 years. Its spectrum changed, too, from that of a classic quasar to a regular galaxy.

Astronomers suspect that all quasars, which were common in the early universe, will eventually transition into humdrum galaxies. The nomenclature “active galactic nuclei” is useful in understanding why. Quasars belong to a larger class of objects called active galactic nuclei, all of which are powered by actively feeding supermassive black holes. Naturally, active galactic nuclei can turn inactive: Over tens of thousands of years black holes run out of gas and dust to eat, so quasars dim and grow quiescent.

There is nothing controversial about the idea that active galactic nuclei can become inactive. What LaMassa and her colleagues doubted was that a quasar could go from active to inactive in just 10 years. Such a dramatic change should occur on a timescale that dwarfs human lifetimes. They looked for different explanations but did not find any that satisfied them. So when LaMassa presented her results at the January 2015 American Astronomical Society meeting in Seattle, it set in motion a rush among the astronomical community to explain her disappearing quasar—and find new ones.

In March Andrea Merloni at the Max Planck Institute for Extraterrestrial Physics in Germany studied LaMassa’s mystery object and suggested that it was not a typical quasar at all. Instead, Merloni proposed, a single star passed near a supermassive black hole and was rapidly torn apart, causing a bright flare that astronomers mistook for a quasar. The light from an event like this would fade over the course of just a few years. Although the explanation matched the timescale, some astronomers found problems with the argument. Julian Krolik, a theorist at Johns Hopkins University who was not involved with either Merloni’s or LaMassa’s studies, points out that a supermassive black hole would not necessarily tear apart a star in the way that Merloni proposed. “The larger-mass black holes can swallow a star whole,” he says, and if that happened, there would be no flare.

LaMassa’s talk in January created enough of a controversy that three additional teams conducted their own research, searching for further examples of these so-called “changing-look quasars.” In September each team posted papers to arXiv, a preprint server used by the astronomy and physics communities. Three teams—one led by Jessie Runnoe of the Pennsylvania State University, one by John Ruan of the University of Washington and another by the University of Edinburgh’s Chelsea MacLeod—discovered a dozen new examples of changing-look quasars. All were found using different methods.

But even with the examples it remains unclear if these objects are shutting down temporarily or for good. An answer might help astronomers better grasp the specifics of how gas and dust flows onto a black hole in the first place and why this process fluctuates over time. Ultimately it could shed light on the processes that helped shape galaxies, including our own.

For his part, Krolik thinks the most likely explanation for changing-look quasars is simple variation. Statistically speaking, it is very unlikely that astronomers just happened to look at these objects, which have lifetimes of billions of years, at the very moments they snuffed out forever. We already know that quasars flicker, varying in brightness by a factor of up to three, Krolik says. A factor of six—although unexplained—is not too far off. And one of the 12 newly discovered objects did not just disappear but reappear. Krolik thinks that this lonely quasar blazed back into existence for the same reason that it flickered out: a variation in the gas and dust flowing onto the black hole.

The next step will be to see if any of the newly discovered dozen brighten over the upcoming years. A few teams would also like to conduct a targeted survey (instead of the serendipitous archival search) to look for more. Although this would likely churn up more fruitful results, the archival search speaks to the fact that there could be interesting things hiding in old images. “I think that's pretty cool that there's all this untapped data out there that could host really awesome discoveries if only we knew what it is to look for,” LaMassa says.