Like a zombie, the Milky Way galaxy may already be dead but it still keeps going.
Our galactic neighbour Andromeda almost certainly expired a few billion years ago, but only recently started showing signs of its demise.
How and why galaxies stop turning gas into new stars is one of the big questions in astrophysics. We may now be on the brink of being able to piece together how it happens. And part of the thanks goes to citizen scientists who combed through millions of images to classify what's out there.
Galaxies grow by making new stars
Like people, galaxies need food. In the case of galaxies, that food is a supply of fresh hydrogen gas from the cosmic web, the filaments and halos of dark matter that make up the largest structures in the universe. As this gas cools and falls into dark matter halos, it turns into discs that cool even further and eventually fragment into stars.
As these stars age and die, they can return some of that gas either via solar winds or by going supernova.
Star formation in galaxies is thus a self-regulated process. The heat from supernova means cosmic gas doesn't cool into new stars as readily, which puts a brake on how many new stars can form.
Most of these star-forming galaxies are disc- or spiral-shaped, like our Milky Way.
But there's another kind of galaxy that has a very different shape. These massive elliptical galaxies are football-shaped. They're not nearly so active " they've lost their supply of gas and have ceased forming new stars. Their stars move on much more unordered orbits, creating the bulkier shape.
Something pretty dramatic must have happened to these elliptical galaxies to produce such profound changes. What?
Blue=young and red=old?
Dividing galaxies into two sorts - star-forming spiral galaxies blazing in the blue light of massive, young and short-lived stars and quiescent ellipticals bathed in the warmer glow of ancient, low-mass stars - goes back to the 20th century.
But once the likes of the Sloan Digital Sky Survey (SDSS) began to record hundreds of thousands of galaxies, objects started emerging that didn't quite fit those two broad categories.
A significant number of red, quiescent galaxies retain roughly a disk shape.
At the same time, blue elliptical galaxies started to surface.
How do these two oddballs " the red spirals and the blue ellipticals " fit into our picture of galaxy evolution?
Send in the citizen scientists
As a graduate student in Oxford, I was looking for some of these oddball galaxies. I was particularly interested in the blue ellipticals.
It became apparent I would need to classify all of the roughly one million galaxies in SDSS with human eyes.
Of course, going through a million galaxies alone wasn't possible.
A short time later, a group of collaborators and I launched galaxyzoo.org and invited members of the public " citizen scientists " to participate in astrophysics research. When you logged on to Galaxy Zoo, you'd be shown an image of a galaxy, a set of possible classifications and a tutorial to help you recognise them.
By the time we stopped recording classifications from a quarter-million people, each of the one million galaxies on Galaxy Zoo had been classified more than 70 times, giving me reliable classifications of galaxy shape, including a measure of uncertainty. Did 65 out of 70 citizen scientists agree this galaxy is elliptical? Good! If there's no agreement, that's information too.
Living in the green valley
The crossroads of galaxy evolution is a place called the "green valley". This may sound scenic, but refers to the population between the blue star-forming galaxies and the red, passively evolving galaxies.
Galaxies with "green" or intermediate colours should be those in which star formation is in the process of turning off, but have some ongoing star formation " indicating the process only shut down a short while ago, perhaps a few hundred million years.
The really exciting moment came when we looked at the rate at which various galaxies were dying. We found the slowly dying ones are the spirals and the rapidly dying ones are the ellipticals.
There must be two fundamentally different evolutionary pathways that lead to death in galaxies.
When we explored these two scenarios " dying slowly, and dying quickly " it became obvious these two pathways have to be tied to the gas supply that fuels star formation in the first place.
Imagine a spiral galaxy like our own Milky Way merrily converting gas to stars as new gas keeps flowing in. Then something happens that turns off that supply of fresh gas. The spiral galaxy is now left with just the gas in its reservoir.
Since these reservoirs can be enormous, our spiral galaxy could go on for quite a while looking "alive" with new stars, while the actual rate of star formation declines over several billion years.
The glacial slowness of using up the gas reservoir means by the time we realise that a galaxy is in terminal decline, the trigger moment occurred billions of years ago.
The Andromeda galaxy, our nearest massive spiral galaxy, is in the green valley: it is a zombie galaxy. It's dead, but keeps on moving, still producing stars, but at a diminished rate compared to what it should if it were still a normal, star-forming galaxy.
Working out whether the Milky Way is in the green valley is much more challenging, as we are in the Milky Way and cannot easily measure it.
But it's entirely possible that the Milky Way galaxy is already a zombie, having died a billion years ago.
Kevin Schawinski is assistant Professor of Galaxy & Black Hole Astrophysics, Swiss Federal Institute of Technology.