On 21 May 2019, from a distance of 7 billion light-years away, our gravitational wave detectors were rocked by the most massive collision yet. From analysis of the signal, astronomers concluded that the detection was the result of two black holes smashing together, weighing in at 66 and 85 times the mass of the Sun respectively.

But what if it was something else? A new study offers a different interpretation of the event. It's possible, according to an international team of astrophysicists, that the two objects were not black holes at all, but mysterious, theoretical objects called boson stars - potentially made up of elusive candidates for dark matter.

The gravitational wave event, called GW 190521, was a spectacular discovery. The object that resulted from the merger of the two objects would have been a black hole at around 142 times the mass of the Sun - within the intermediate mass range that no black hole had ever been detected before, called the black hole upper mass gap.

That was extremely neat, but there was a huge puzzle - the 85 solar-mass black hole allegedly involved in the collision. According to our models, black holes over about 65 solar masses can't form from a single star, like stellar mass black holes.

That's because the precursor stars that would produce a black hole in this mass range are so massive that their supernovae - known as pair-instability supernovae - ought to completely obliterate the stellar core, leaving nothing behind that could gravitationally collapse into a black hole.

While our understanding of the formation of stars as 'twins' doesn't neatly allow for pairs of stellar black holes to be born close enough to combine, it's likely that the explanation is two smaller black holes merging. But if we go by the data alone, another model fits even better.

It's possible that the black hole was the product of an earlier merger between two smaller black holes. Led by Juan Calderón Bustillo of the Galician Institute of High Energy Physics in Spain, the research team has determined boson stars would be a perfect match for the numbers.

"Our results show that the two scenarios are almost indistinguishable given the data, although the exotic boson star hypothesis is slightly preferred," said astrophysicist José Font of the University of Valencia in Spain.

"This is very exciting, since our boson-star model is, as of now, very limited, and subject to major improvements. A more evolved model may lead to even larger evidence for this scenario and would also allow us to study previous gravitational-wave observations under the boson-star merger assumption."

Boson stars are, at the moment, purely theoretical, and have never been detected before, but they are of increasing interest to astronomers, particularly in the search for dark matter.

They are, like black holes, predicted by general relativity, and are able to grow to millions of solar masses at a very compact size.

As we have previously reported, where stars are primarily made up of particles called fermions - protons, neutrons, electrons, the stuff that forms more substantial parts of our Universe - boson stars would be made up entirely of bosons. These particles - including photons, gluons and the famous Higgs boson - don't follow the same physical rules as fermions.

Fermions are subject to the Pauli exclusion principle, which means you can't have two or more particles with the exact same quantum states, which includes the space they sit in. Bosons, however, can be superimposed; when they come together, they act like one big particle or matter wave. We know this, because it's been done in a lab, producing what we call a Bose-Einstein condensate.

In the case of boson stars, the particles can be squeezed into a space which can be described with distinct values, or points on a scale. Given the right kind of bosons in the right arrangements, this 'scalar field' could fall into a relatively stable arrangement.

Boson stars might actually look a lot like black holes, except for one characteristic: they don't have an absorbing surface that would stop photons, or an event horizon, so they would appear totally transparent. They're basically compact blobs of Bose-Einstein condensate in space.

The countless particles making up such massive stars would ironically need to be incredibly light, with millions of times less mass than an electrons.

Interestingly, this kind of ultralight boson would also be a candidate for dark matter - the unknown, unseen mass responsible for all the extra gravity floating around the Universe that we can't account for. So finding boson stars would go at least some way towards solving one of the biggest mysteries of the cosmos.

According to the team's calculations, if GW 190521 was a merger between two boson stars, the masses and distances involved would be different, but it would solve the problem of that 85-solar-mass black hole.

"First, we would not be talking about colliding black holes anymore, which eliminates the issue of dealing with a 'forbidden' black hole," Calderón Bustillo said.

"Second, because boson star mergers are much weaker, we infer a much closer distance than the one estimated by LIGO and Virgo. This leads to a much larger mass for the final black hole, of about 250 solar masses, so the fact that we have witnessed the formation of an intermediate-mass black hole remains true."

In the team's scenario, when the two boson stars collided, they formed a larger boson star that could have become unstable and collapses down into a black hole, so it's actually impossible to tell whether the boson star interpretation is correct, even if we could see it clearly across the revised 1.9 billion-light-year distance.

Instead, the analysis gives us the tools for studying intermediate-mass gravitational wave events going forward in the context of boson stars as well as black holes, with the hope of finding answers in the future.

"If confirmed by subsequent analysis of this and other gravitational-wave observations," said astrophysicist Carlos Herdeiro of the University of Aveiro in Portugal, "our result would provide the first observational evidence for a long-sought dark matter candidate."

The research has been published in Physical Review Letters.