star cluster. There were maybe twenty or so stars, of roughly
equal brightness, with hints at the edge of visibility that many
more existed. Surprised I’d never heard of such a glorious cluster
close to a tourist stop as popular as the Orion Nebula, I looked it
up in my star atlas and found, where the cluster should be, noth-
ing but a blank space. In those days before the internet was avail-
able at home at the flick of a screen there was no way for me to
immediately investigate further, and I proudly picked up a pencil
(not a pen—I obviously wasn’t quite that confident!) and marked
its position very carefully on the atlas, next to a carefully calli-
graphed label: ‘Lintott 1’.
I went to bed that night not really believing that I’d made a dis-
covery, but I couldn’t stop a tiny part of my brain thinking that
there was at least a chance that I might have done. Of course,
when I made it to the internet* the next day the cluster turned
out to be well known. Its true name, NGC 1981, doesn’t have the
sonorous ring of Lintott 1, but has the advantage of having been
in use by astronomers since the late nineteenth century (Plate 1).
Minor though it was, the incident stuck with me as my closest
contact with a long-ago epoch when anyone with time, luck, and
modest equipment could make a discovery. The astronomical
books and magazines I immersed myself in were full of tales of
amateurs discovering enormous storms on Saturn, spectacular
comets, or the twinkling of variable stars, but they did so partly
because they had kit beyond my wildest dreams. People like
* People younger than me may be surprised by the idea that the internet wasn’t immediately available from home. It was a bleak time.
8 How Science iS Done
retired telecoms engineer Tom Boles of Suffolk brought enor-
mous dedication to their searches, in Tom’s case resulting in the
discovery of more than a hundred supernovae (explosions that
mark the death of massive stars), but they also depended on tens
of thousands of pounds’ worth of automated telescopic power.
These super-amateurs were making spectacular progress, but
they were doing so by effectively professionalizing their astron-
omy, often spending a fortune on advanced kit. The market
for buckets and spades being unlikely to expand to a sufficient
extent, I felt that finding even something as quotidian as an
undiscovered asteroid—once so commonplace as to have
attracted the reputation of celestial vermin—would be forever
beyond my reach. My romantic vision of amateur astronomy
with a small telescope dead, I moved on to become a professional
astronomer.*
Here, as with the amateurs, I found a community and a way of
life that was under severe assault from the onrushing forces of
technological advance. While once individual universities or
particular countries aspired to have their own research-grade
telescope, a severe epidemic of ‘aperture fever’—the desire for
larger and larger telescopes—has hit the astronomers of the
world, leaving us with no option but to pool resources in ever
larger and more expensive facilities. For most of the twentieth
century, the world’s largest working optical telescope was the
200-inch Hale Telescope at Mount Palomar in California. Built
* That’s not to say that there isn’t anything worth doing with a small telescope. I’ve had hours of pleasure staring at the Moon, Jupiter, the Orion Nebula, and countless other objects. If you’ve never looked through a telescope, find your nearest astronomical society right now, go along to a public observing night, and get them to show you the sights. It’s more than worth it, even if you don’t get a comet named after you. There really is something special about having photons from far away objects hit your retina, something which just can’t be replicated by looking at even the most impressive photos.
How Science iS Done 9
during the 1930s and then commissioned in the immediate after-
math of the Second World War, the 200” is a beautiful relic of a
bygone age, to the extent that the original design’s reliance on a
newfangled technology—welding—was considered risky when
it was completed. The welding worked, and the telescope was
unrivalled until the 1990s.*
Since then, a host of telescopes with mirrors eight metres
(307 inches) or more across have been made, and astronomers
are working on three separate projects that take us up to the
equivalent light-collecting area of a mirror thirty metres or
more across. One of the projects, now known logically enough
as the European Extremely Large Telescope (EELT), started life
as a concept called OWL—the OverWhelmingly Large Telescope.
That was supposed to sport a mirror one hundred metres across,
and it’s been suggested they could have kept the acronym but
have it stand for Originally Was Larger. The EELT will still be a
monster. Its dome, for example, is the size of a sports stadium,
but competition to use such an enormous facility when it’s
completed in the mid-2020s will be intense. With fewer large
facilities available in each generation to be shared among the
growing community of astronomers, the pressure to make the
best use of every second of observing time has grown, and that’s
had serious consequences.
My PhD work involved several trips all the way from University
College London in the heart of Bloomsbury to the Big Island of
* A larger telescope was built by the Soviet Union in the mid-1970s, but it was plagued by problems bad enough for most to discount it in the competition for the world title. The original mirror was so bad it had eventually to be replaced, but even then the telescope remained almost neurotically sensitive to changes in temperature. That being said, with a lot of help from those who know it well, my collaborators and I have managed to get good data from it, so perhaps writing it out of the history books is a tad unfair.
10 How Science iS Done
Figure 3 The summit of Mauna Kea in Hawai’i, home to some of the best conditions for astronomical observing anywhere in the world.
Hawai’i, with the aim of taking advantage of the crystal clear
skies available to observatories on top of the (hopefully) dor-
mant Mauna Kea volcano (Figure 3). It is a spectacular and other-
worldly place, an hour or two’s drive from palm-lined beaches
yet regularly covered in snow. It is, in fact, the summit of the
world’s tallest mountain, double the height of Everest when
measured from its summit more than 10,000 metres above sea
level to its base deep in the ocean, and high enough that the
low air pressure at the top presents a serious obstacle to clear
thinking—an environment so hostile that the only thing that
lives on the summit is a species of beetle found nowhere else on
Earth. I’ve always wanted to see a Wēkiu bug, as they’re known—
a strange creature which sits and waits for food in the form of
smaller insects to be blown up to it by the winds that sometimes
sweep across the summit, and is surely evolving to consume
astronomers—but they are very rare.
Mauna Kea must be a strange place to spend most of you
r days.
I wasn’t too surprised to discover that several of the staff had been
How Science iS Done 11
submariners in their previous life, and were therefore presumably
acclimatized to spending lots of time shut into metal boxes in a
dark and hostile environment.
Whatever it’s like to spend a significant proportion of your
time there, it was a glorious place to visit, and I loved leaving the observatory buildings for a glimpse of a Mars-red dusty landscape high above bright blue sea or, better, for a spectacular
night sky illuminated only by the stars and, just occasionally
and far below, the flickering that marks the flow of magma from
the still-active Kı̄lauea volcano into the ocean. (Astronomers
have had a slightly more exciting time with Kı̄lauea recently, as
the major eruption in the summer of 2018 produced a cone of
sparks that could be seen from the summit as well as producing
warnings of shut airspace and toxic vog, a volcanic fog, across
the island. I know at least one survey that’s six months behind
because of delay caused by fear of volcano-induced earth-
quakes.)
The dream of travelling to this amazing place, and being there
on a scientific mission with purpose, was a large part of what
attracted me to being an astronomer, but things are changing
rapidly. To gain access to the telescope that I used in Hawai’i, one
bids for access, competing against other astronomers to explain
the (scientific) justification for being allowed to take the controls.
This can be tricky, requiring one simultaneously to argue that
the results will be transformative, and that you know what they
will be sufficiently well that you can promise the time won’t be
wasted.
Still, if successful then the telescope was yours for the night
and all you needed was a bit of luck and a clear sky. If conditions
were lousy (and too much time spent under beautiful Hawai’ian
skies tends to lead to over-optimistic planning or some very
picky observers who want nothing but the calmest, stillest
12 How Science iS Done
nights) then there is nothing for it but to turn round and apply
again. Desperation, not to say depression, could set in; I remem-
ber observing with a radio telescope while rain swept across the
mountain top, washing away any chance of decent data, and on
another occasion nearly a week spent kicking pebbles along the
gloriously palm-swept beach in high dudgeon at the sheer unfair-
ness of cloud.
Worse from the observatory’s point of view are the nights
where I got lucky, or at least luckier than I should have. My PhD
was mostly on the subject of astrochemistry, and I’d got time to
go looking for molecules in space using a dish fifteen metres
across known as the James Clark Maxwell Telescope, or JCMT.
The JCMT works in the region of the spectrum that astronomers
call the ‘sub-mm’ which and everyone else calls microwaves,
which explains why it was high on the summit of a volcano. Your
microwave oven works by firing waves at a wavelength chosen to
excite the water in your food; the water in the atmosphere emits
radiation at similar wavelengths, so from ground level the micro-
wave sky shines brightly. By climbing a mountain we could get
above most of the atmosphere’s water and see clearly out into the
cosmos.
I was searching for chemicals like hydrogen cyanide in and
among the clouds of star-forming regions like the Orion Nebula.
That isn’t quite as quixotic a quest as it sounds—the surfaces of
dust grains in star-forming nebulae provide excellent sites for the
sort of chemistry that forms complex molecules, and observing
them provides more information than physical measurements
alone ever could—but because for the most part we were happy
with a single detection, rather than needing a map, a blurry view
would do and these observations could be completed when con-
ditions weren’t great. Nights with the best ‘seeing’—those when
the air is crisp and still—would be better spent on high- resolution
How Science iS Done 13
mapping or imaging projects, but if one of those came along
when I was on the telescope then it was going to be used for
detecting hydrogen cyanide no matter how good the skies
became.
When both the astronomer sitting miserably on the beach and
the observatory management started wondering how to make
their multimillion dollar instruments more productive, it was
only a matter of time before we found our way to a more efficient
means of doing things. Gradually, over the past twenty years, the
major observatories have shifted their operations to a model in
which observations are dynamically scheduled; if the conditions
improve, then high-priority objects which require the best wea-
ther can be targeted and previously scheduled work shelved tem-
porarily. In most cases, it’s become rare to travel to a telescope,
and rather than carrying back precious images in triumph from
Hawai’i, or Chile, or the Canary Islands, the arrival of fresh data
is signified by the ping of an email hitting my inbox. I miss my
observing trips, but I worry more about the next generation of
students who won’t necessarily have any hands-on experience
with where their data come from.
A healthy respect for data is critical in developing scientific
scepticism, but once the astronomer was removed from the pro-
cess of observing it at least became clear that new ways of work-
ing were possible. Whereas astronomers like me are used to
observatories, general purpose facilities that are built for a multi-
tude of tasks, what we’re getting increasingly in astronomy are
experiments; there’s a tendency to move away from targeted
observations altogether and towards collaborative surveys of
large chunks of sky. Such surveys produce ever-larger reposito-
ries of data held in trust for hundreds or even thousands of scien-
tists to use, none of whom need have gone anywhere near the
telescope upon which their research depends.
14 How Science iS Done
The most successful by far of these projects is called the Sloan
Digital Sky Survey. It’s named for and funded by the foundation
set up in the memory of Alfred P. Sloan, the man whose modern
and, above all, efficient management techniques turned General
Motors into the colossus that bestrode mid-twentieth-century
America. He would have approved, I suspect, of the business-like
way that his namesake fulfilled its primary mission, scanning the
sky from its home in New Mexico in order to create a three-
dimensional map of the Universe including nearly a million
nearby galaxies (Figure 4).
Is a million galaxies a lot? A single, medium-sized galaxy like
our own Milky Way contains roughly one hundred billion stars,
and so Sloan’s sample contains plenty of interest. Yet there are at
Figure 4 The Sloan Digital Sky Survey telescope in New Mexico. The
main mirror is 2.4 met
res across.
How Science iS Done 15
least one hundred billion galaxies or so within the span of the
observable Universe—as many galaxies, in fact, as there are stars
in the Milky Way—and so this, the most detailed map we have of
our surroundings, only includes 0.001 per cent of the available
sample (Plate 2). Even by the standards of Earth’s medieval map-
makers, that’s a lot of undiscovered country marked ‘Here be
dragons’.
Yet the effort isn’t hopeless, and another analogy might help.
Election pollsters in the US are faced with predicting the behav-
iour of something like a hundred million people, yet typically
use a sample of maybe a few thousand voters who they can reach
on the phone or online. The ratio between observed and esti-
mated samples is almost the same, and so we might conclude
that we should expect our knowledge about the Universe based
on the Sloan observations to be about as accurate as a single
opinion poll. Actually, things aren’t that bad. Not only are galax-
ies much simpler than people (a fundamental truth that makes
astrophysics possible), but the Universe is much less variable on
large scales than is opinion in America. For most purposes,
therefore, we can assume that the volume covered by Sloan is a
typical chunk of the Universe, and draw conclusions about the
whole based on what the survey shows us.
One can, of course try and look deeper into the Universe as
well. One of the great advantages astronomers have in trying to
understand the Universe is our ability to see into the past. Sloan
is a survey of the local and thus the present-day Universe, but as
we look further away we receive light which has taken billions of
years to reach us. By combining our views of near and far we can
piece together the whole story.
The main scientific goal for which Sloan was built was to study
the carefully plotted positions of its million galaxies in order to
measure the expansion of the Universe. The galaxies are, in this
16 How Science iS Done
kind of study, nothing but trace particles, carried along on the
grand expansion of space like so much flotsam carried by a river
in flood. The first step towards this grander goal, though, was to
identify enough galaxies, and that meant imaging a large area of
The Crowd and the Cosmos: Adventures in the Zooniverse Page 3