The Crowd and the Cosmos: Adventures in the Zooniverse

by Lintott, Chris

  ‘Could we be aided in this matter by the cooperation of a goodly

  number of amateurs, ‘we would perhaps in a few years be able

  to discover laws in these apparent irregularities, and then in

  a short time accomplish more than in all the 60 years which

  have passed since their discovery. I have one request, which is

  this, that the observations shall be made known each year.

  Observations buried in a desk are no observations. Should they

  be entrusted to me for reduction, or even publication, I will

  undertake it with joy and thanks, and will also answer all

  questions with care and with the greatest pleasure.’*

  It is a fabulous call to arms—‘observations buried in a desk are

  no observations’ would be a great motto for some society or

  other.† I love the sense of a deal being struck between those

  taking the observations and Argelander himself. On the one

  hand, we have the (presumably unfunded) volunteer with their

  telescope. On the other, an eminent professional scientist. Data

  can be passed from the former to the latter—but only if

  Argelander too puts his back into it and makes use of the data.

  * Translation by Annie Jump Cannon in Popular Astronomy, 1912, from an original in the Astronomisches Jahrbuch of 1844.

  † To my mind, greatly preferably to the Royal Astronomical Society’s motto, adopted from Herschel: ‘quicquid nitet notandum’, or ‘whatever shines, let it be observed’. Science teaches us that the real work is only beginning when observations are written down. The American Astronomical Society has a mission statement, not a motto.

  No Such ThiNg aS a New ide a 99

  Oh, and part of the deal is that he has to communicate results

  and answer questions from his observers.

  It seems important that Argelander is offering more than a

  one­way exchange. As far as I know this is the first example of a

  professional scientist so explicitly writing about the give and

  take of this way of collaborating to get science done. As Galaxy

  Zoo took off, I certainly felt the obligation to try and respond to

  questions, though I can’t claim to have always faced the task ‘with

  the greatest of pleasure’. What is also reflected in Argelander’s

  work is a somewhat formal division of responsibility; observation can be safely distributed, but analysis is specialized and central. One can argue about which is primary (and whether Darwin was being deliberately or falsely modest when referring to himself as a mere ‘compiler’), but there is a settled order here.

  This way of organizing things was effective, and it enabled

  Argelander to work on a scale that was inaccessible to astronomers

  of previous generations. The catalogue Argelander and colleagues put together contained the details of more than 300,000

  stars, and was the definitive work of pre­photographic stellar

  astronomy, at least for the northern hemisphere. It remained in

  use for years, and his categorization of variable stars remains the

  standard today. If you visit the astronomy facilities in Bonn,

  you’ll find that in 2006 they were renamed the ‘Argelander

  Institute’ in his honour; a recognition, I’d like to think, of the

  power of asking for help.

  Networks of amateur astronomers survive too. Data on stellar

  variability, especially on timescales of decades or more, depend

  on the catalogues assembled by the American Association of

  Variable Star Observers, an organization with worldwide reach

  whose observers have assembled more than twenty million

  records since its founding in 1911. Rainfall observers may not,

  these days, form extensive networks but the Audubon Society’s

  100 No Such ThiNg aS a New ide a

  Christmas Bird Count is still going strong. This annual birdwatching festival has been operating since 1900, having been introduced partly for scientific interest and partly as an alternative to the then common tradition of marking the holiday with competitive hunting. In the UK, biological recording of the presence or absence of species depends on a network of societies, many of them dating from the late nineteenth or early twentieth

  century, with specialisms ranging from orchids to the British

  Pteridological Society (ferns, since you ask).

  Twenty­first­century researchers are fond of pontificating

  about the problems caused by big data, the sudden flood of digital information among which we struggle to pick out signals of interest. Yet the appearance of modern instruments and vast networks of the kind described above caused an earlier deluge of data, and brought a very different set of people into the scientific

  enterprise. These were the first ‘computers’, people rather than

  machines, and they soon accounted for the majority of staff

  employed at observatories.

  The job of ‘computer’ was established at the Royal Observatory

  in Greenwich, for example, as early as 1836, and survived until

  1937, a little more than a century. Their arrival broke the tradition by which it was the Astronomer Royal and his specialist assistants

  who did the work to make their own observations useful, and the

  staff quickly grew. The original computers, working eight hours

  or more a day at tedious and repetitive calculations, were

  recruited by looking for poor but bright students from local

  schools. However, as the century wore on it became clear that

  this was nothing more than a stopgap solution; the wages were

  abysmal and the work tedious, and with little prospect of promotion most computers moved on. By 1890 the Greenwich staff had hit on the idea of solving this by employing women who had

  university experience; such staff would be skilled enough to

  No Such ThiNg aS a New ide a 101

  work as more than a mere calculator, but would lack other

  opportunities for scientific participation.

  The papers which record the decision to employ women

  explicitly make clear that it was felt that Greenwich could take

  advantage, attracting to this often menial job women whose scientific opportunities would otherwise be lacking. If that was the marketing scheme, it was not a success. Though women continued

  to be employed at Greenwich, opportunities for promotion were

  gradually opened up for men, and those with qualifications

  (most often a degree) began to take up what had once been junior

  and menial posts, seeing them as a stepping stone to higher

  things. The women were once again squeezed out of even this

  small foothold in the scientific enterprise.

  For a time, though, the position of these functionaries allowed

  a different sort of engagement with scientific data. What sorts of

  jobs were these human computers undertaking? Most of the

  work at Greenwich was positional astronomy, and results would

  be recorded in the form of measurements which were straight

  from the telescope, perhaps as a distance between two stars.

  These would have to be converted to some standard reference

  frame, and celestial coordinates assigned. Systematic effects like

  the influence of the Earths’ atmosphere, which varies with the

  height of a source in the sky, must be accounted for. Even once

  that’s done, single observations of a typical star are hardly going

  to carry much information, and catalogues must be compiled

  and cross­checked, and global properties derived.

  These c
alculations are the very stuff of which science is made,

  but just as with Argelander and his observers we see in the existence of the computers a division of responsibility. Observers—

  whether employees or volunteers—provide data. Computers do

  the processing, turning tables of data into results; the two are

  even separated by time, with observers producing data during

  102 No Such ThiNg aS a New ide a

  the night that can be processed by computers during the day,

  before being studied by scientists who publish their results. Each

  subsequent stage of analysis depends on the former—indeed, it

  could hardly exist without it—but only the later stages are visible

  to the wider world. We celebrate the scientist who interprets the

  observation, not those who made it possible.

  None of this is news, at least not to the accomplished historian

  or sociologist of science. You, my sophisticated reader, don’t

  need me to tell that the real story of how science has proceeded

  over the centuries is more complex than the standard procession

  of dead, white, bearded men with theories might suggest, and

  this chapter hasn’t tried to do more than offer a potted set of

  anecdotes. These stories do, however, illustrate that right back at

  the time when our modern notion of what it meant to be a scientist was being established—when we had a much more fluid idea of what science was than has been the case for most of the last

  century—it is possible to trace disputes about status, about the

  correct division of work between the classes of those involved in

  research. Back in the twenty­first century, we set up Galaxy Zoo

  to get work done. It soon became apparent that the real power

  and interest of the project lay in thinking about precisely these

  issues, and that began with volunteers doing more than just

  clicking on buttons to classify galaxies.

  4

  INTO THE ZOONIVERSE

  Looking back at the early days of Galaxy Zoo with more than a

  decade’s perspective, it seems to me to be a strange and

  marvellous thing, this idea that so many people would give up

  their time to collectively contribute to science. Occasional critics

  carp that classifying a few galaxies isn’t participating in science—

  that the claim to have done science should be reserved for those

  who designed and set up the project, and who interpret the

  results.

  As I said in the last chapter, I’m less dogmatic. Any scientific

  project rests on the contributions of many people, whether it’s

  those who operate the telescope up on a lonely mountain top or

  people like me, whose daily life is much more likely to involve

  emails and admin than a ‘Eureka’ moment. I know the Galaxy

  Zoo crowd have done science, because there’s an ever-growing

  pile of academic papers with new scientific results within that

  wouldn’t have existed without them.

  Better still, the ideas in those papers have been adopted and

  echoed by the rest of the community. We were even thrilled

  when people started to use our results without pausing in their

  texts to dwell on ‘citizen science’, taking it as a sure sign that we were producing data of a high-enough quality that authors

  104 Into the ZoonIverse

  didn’t feel the need to justify or explain their use of it. (Things

  took a slightly odder turn when two philosophers wrote a

  paper which, while calling for ‘a sociotechnological turn in the

  philosophy of science’, which I’m afraid to say you’ll have to

  read about elsewhere, compared the rate at which Zooniverse

  papers were cited to others using the same data. Apparently we’re

  as a group as productive as a world-leading research institute.

  Nice to know!)

  The downside is that, all this time later, it’s rather difficult to

  briefly summarize what we’ve found. Galaxy formation is messy,

  and that messiness—the fact that many different things control

  how galaxies first form and then change over billions of years—

  makes a nice, clean story hard to find, at least for now. That’s

  what science is like sometimes, even if it makes writing a book

  chapter harder. So, instead of trying to present a comprehensive

  view, let me tell you a couple of stories that will give you an idea

  of the kind of thing we’ve been able to do with the results from

  Galaxy Zoo.

  One problem we’ve tried to attack is to try and understand

  what happens when two galaxies collide with each other. Merging

  like this certainly seems important. The early Universe was filled

  with scrawny protogalaxies, each less than a hundredth the size

  of a typical galaxy today, and these seem to have, over the long

  span of cosmic history, gradually collided and merged to form

  larger and larger systems. This process isn’t finished yet—the

  grand collision of the Milky Way with Andromeda that I men-

  tioned earlier isn’t due to happen for another four or five billion

  years’ time, but when it does happen, our computer simulations

  make it clear that it’s likely to be a spectacularly messy and dis-

  ruptive event.

  When two large galaxies like these collide with each other, a

  cosmic ballet ensues. The first approach sees the galaxies fly past

  Into the ZoonIverse 105

  each other, their mutual gravitational attraction distorting their

  previously neat discs and creating long streams of stars (Plate 6).

  These are tidal tails, unstable creations of the merging process,

  and as they begin to fall back towards the main body of each gal-

  axy the two discs turn and plunge back together once more.

  This repeated encounter creates new distortions, and further

  disruption as the merging system takes on a wide variety of

  forms. We can see this stage of the process in nearby galaxies,

  whose names conjure up appropriate images—the ‘Antennae’,

  with two long streams stretching away from a bifurcated body

  the ‘Mice’, imaged beautifully with the Hubble Space Telescope, with long tails revealing a recent interaction. Such a stream has even

  been spotted stretching between Andromeda itself and the third

  large member of our Local Group, M33.

  Apart from being flung out of orbit, stars which formed before

  the merger will continue as they were before; even within a gal-

  axy of a hundred billion stars like the Milky Way, there is enough

  space in space to make a collision between two stars during a

  merger vanishingly unlikely.

  That’s not to say we shouldn’t expect fireworks when the Milky

  Way and Andromeda collide. Gas clouds do collide with each

  other and the result is a spectacular boom of star formation. The

  Earth may not be the best place to watch, as the Sun will by then

  have entered its red giant phase and swallowed our home,* but if

  you can make it to a suitable planet then you should expect a

  spectacular night sky, speckled with newly formed and brilliant,

  * I may be being unfair to the Earth’s prospects as a long-term observing platform. As the Sun converts hydrogen to helium it loses mass, and, because of the law of conservation of momentum, our planet spiral
s slightly outwards. There is therefore some chance that the Earth may survive the Sun’s swelling into a red giant, though how reassuring you find the chance of our planet’s future existence as a charred cinder is perhaps a matter of personal taste.

  106 Into the ZoonIverse

  massive stars. The view from inside what’s called a starburst

  galaxy must be absolutely wonderful.*

  Though such a spectacular rate of star formation most likely

  can’t be sustained, the long-term effect of a merger might be to

  change the shape of the colliding galaxies for ever. In the case of

  the collision between the Milky Way and Andromeda—two

  discs—the result according to most simulations is their trans-

  formation into an elliptical.

  This makes a certain amount of intuitive sense; discs are

  ordered systems, their stars orbiting in concert around their

  centre, and a serious disruption will see stars kicked up out of the

  disc and into the more random pattern of movement which char-

  acterizes ellipticals. A new, unified galaxy is produced (what

  some researchers insist, despite everything, on calling Milkomeda

  or—hardly better—Milkdromeda), larger and more massive

  than before and ready to continue life as a stereotypical elliptical.

  Perhaps the last stage of such an event takes place deep

  inside the new galaxy, at its core, as the supermassive black

  holes that previously inhabited the centres of the constituent

  galaxies dance slowly around each other, losing energy in the

  form of gravitational waves and spiralling inwards, eventually

  merging. A small number of galaxies are known that have

  double or even triple black holes at their centres; though they

  look otherwise undisturbed, these are most likely the products

  of recent mergers.

  Galaxy Zoo must contain many such galaxies, observed a

  few billion years after the end of the merger. Can we tell, just by

  looking at the galaxy, that anything spectacular had happened?

  * Of course, the odds of your planet being blasted with lethal radiation from a nearby supernova is greatly increased in such a system. One can’t have everything.

  Into the ZoonIverse 107

  The tidal tails of stars flung from the centre of the system will