Science is creative… obviously

Science is creative. I don’t mean in the “I made a mood board and have you seen my new beret?” way, although honestly some scientists do dress like art teachers so who knows, but in the real, difficult, frustrating but awesome way where you’re trying to understand something that doesn’t want to be understood, using tools that only sort of work, in systems that are probably lying to you, while your cells die, your assay fails, your supervisor asks if you’ve tried just repeating it and somewhere in the back of your mind a small imposter-syndrome-gremlin whispers, “What if you’re just a bit shit at this?”

So, er, yeah. Creative.

I wrote a version of this yonks ago and then I read a paper this morning and recently I was chatting with some lovely people who work in design. They were a mix of illustrators, graphic designers and even a former fine artist, which is to say people who are legitimately allowed to call themselves Creatives. The conversation eventually moved on to creativity, what it is, who has it, can you learn it and whether owning a MacBook makes you spiritually closer to it (doubly so if you’re in a coffee shop), and as former scientist I offered the point that working at the bench also needs a fair bit of creativity.

Shock and not-a-little scepticism are what that remark was met with.

“Really? Science? Science is creative? Isn’t science just following steps or doing something with data?”

No. No it bloody isn’t.

I mean, well, yes, sometimes it is following steps and doing something with data, obviously. You can’t just waltz into a lab, whip out a pipette, shout “Creativity, ON!” and expect nature to reveal her secrets. There are methods and protocols and controls and spreadsheets with names like cell_analysis_final_CSedits_revised_AKedits_v7_ACTUALLY_THIS_ONE.xlsx because scientists are still human and therefore terrible at file management.

But no, science isn’t just data and following steps. It’s like saying writing is just pressing keys or parenting is just keeping a child broadly alive until adulthood, which, okay, some days, yes, but also no.

Biology is a beautiful bastard

Science, particularly biology, is a glorious smorgasbord of processes and pathways, networks and cycles, proteins and peptides and everything overlaps somewhere at some point because biology has absolutely no respect for our need to draw clean diagrams. Physics and chemistry have this as well, I’m sure, but I don’t know them well enough so I’m not going to comment on the lesser sciences (jk).

The point is, it’s a bloody mess. And we love it. Actually that’s one of the many appeals of even studying biology: untangling, or at least taking a glimpse into, that underlying complexity before it wriggles away again like a smug little eel. You look at a living system and you think, “Right, okay, so this gene affects that pathway, which affects this cell state, which affects that tissue structure, which then feeds back into the original signal… Oh, but only on Tuesdays and only if the temperature is exactly within plus or minus 0.3 degrees C and that one batch of serum hasn’t decided to become evil.”

It’s beautiful and infuriating but mostly beautiful. I mean, we’ve got gene regulatory networks with complex time-dependent feedback mechanisms at play during development. We’re overflowing with cascades of cytotoxic events triggered by a host of exogenous compounds. There’s the dynamic interplay of ions and electrically responsive channels working together to move signals along neurons at incredible speeds. And we have myriad other phenomena that work with and against one another to achieve a balanced outcome within all living organisms that, for the most part, go completely unnoticed as you bumble around your day.

If we want to understand what’s behind that beautiful complexity, we need some spark of creative insight to get us started. Because contrary to what you’ve seen in the movies, science doesn’t begin with certainty. It begins with looking at something strange and thinking, “Hang on, what the crap is that?”

Curiosity is where the trouble starts

Observation and curiosity is where creativity in science gets going. We observe the phenomena and we start to think about them. It isn’t always terribly academic; we might just wonder to ourselves, “Err, WTF just happened?!” but it’s a definite start. Not to be too schmaltzy but what you’re hoping for at this is stage is a little dash of feeling wonderstruck and a desire to know more about something unknown.

Okay, so curiosity is our driver here, we want to know how and why something occurs, so we turn to hypotheses and experiments. We form an initial hypothesis based on a degree of intuition and, of course, a fair bit of research into what others have done and we start to build a mental picture of how something might occur. And then we try to disprove it. Yeah, that’s the bit people miss. This is called the null hypothesis and I’m not going into it here but experimental design could be a whole other blog and I don’t have time, damn it!

Science isn’t “I reckon this is true, so now I shall prove it and receive applause”. Science is more like, “I reckon this might be true, but because I am committed to emotional damage as a career structure, I will now design a series of experiments specifically intended to kick the legs out from under my own idea.”

And dude, that’s creative as hell. We put hammer and nail – or pipette and microscope – to our assumptions. Creativity takes charge here because to gain even a modicum of insight into what is truly happening, we have to design clever, creative experiments. This means controlling for reagents and unwanted effects, sure, but also controlling for our own biases.

It takes a truly creative thinker to devise a series of experiments that not only probe the nature of something, but which also successfully rule out potentially confounding variables – including from ourselves – that might distract us from the truth.

And yes, being wrong is part of that, a “My hypothesis has died and I now have to explain this to people who funded me” kind of wrong, which in turn forces you to get even more creative (not just with explanations for your failure) because you need to think really hard of better ways to figure what on earth is happening in your system.

Thinking sideways

Speaking of thinking, after the experiments comes the analysis. If experiments are a test of creativity, then looking at the data is a test of lateral thinking and resolve. Data are reflective of the science and so more often than not that means they’re messy. The first step here then is making the data more tidy and interpretable.

Like cleaning your house, cleaning and tidying data is rarely a calm affair (everyone rage hoovers, right?). Now’s the time your mates find you hunched over a laptop, trying to work out whether the thing you’re seeing is biology, technical noise, the wrath of the gods or the consequence of naming your samples something useless like “sample 1” because past-you was an arrogant plonker who assumed future-you would remember what anything meant.

The challenge can come from the sheer number of ways to modify and analyse data. As scientists, we need to be able to look at large and complex data sets, think about the many possibilities available and choose the most appropriate ones. Sorry, i can’t make that bit sound cool.

Okay, data are clean and tidy, now you actually look at your results. Interpreting data demands lateral thinking, and we won’t be relying on simple “because x then y” because while correlations are wonderful initial indicators of links between results, they’re also red herrings that lure us to a false conclusion of causation. Plus we’re not troglodytes, darling, for if science has one eternal enemy it’s probably This line goes up when that line goes up so obviously I have solved the universe. No, Gary, you have not.

So we have to see beyond that. We have to look at broader, less obvious possibilities. We have to think about and create additional layers of robust analysis to tease apart and link together seemingly disparate results.

This is where science gets properly interesting because you’ve moved from processing information to trying to get all up inside it, turning it over, wondering what happens if you plot it differently, squint at it differently, compare it to something from another field or stop asking the question everyone else asked because maybe the question itself is the problem.

The Royal Society paper I’ve been reading puts this nicely, saying how science needs both vertical and lateral thinking (1). Vertical thinking is the logical step-by-step stuff, the testing, the sequence, the rigour. Whereas lateral thinking generates and rearranges ideas and challenges assumptions. So you need both. If you only have vertical thinking and you get science as a production line. Very tidy. Very measurable. Very dead behind the eyes. If you only have lateral thinking you get a hairy bloke in a linen shirt on instagram explaining quantum healing at his wellness retreat and nobody wants that. Well, some people do, but they also think celery juice is great, so ignore them.

In the end, you’ve seen a cool but unexplained phenomenon in nature. You’ve wondered what on earth is going and come up with a rough hypothesis, a rough idea. Then you’ve meticulously created a series of experiments to figure out why that might be happening and controlled for all the variables along the way. You’ve then cleaned and analysed a bunch of data, looked way too closely for way longer than you’d like to admit and found a weird result, which prompted more experiments – and controls – to figure out what that little weird bit means. At each step you’ve been presented with completely unknown outcomes and had to think creatively about not only how it might be occurring, but how you might sensibly and reproducibly test it. It’s a whole, never ending cycle of observe, create, test, analyse, create, observe, analyse, cry, test, create… ad infinitum. Then all you need to do is write it all up!

The system is squeezing the weird out

Science is creative, yes. Scientists are creative, also yes. But the system science lives inside isn’t always built for creativity. Modern science has shifted towards a production model where narrow measures of productivity and excellence get rewarded; papers, grants, outputs, metrics, impact, economic value, institutional prestige, all the usual beige machinery of Please prove your worth using a table no one wants to read.

And yes, obviously some of that matters. Scientist need funding and funders need to know people aren’t just spending public money on coffee and a suspicious number of conferences in warm countries. But when the system only rewards what can be counted, it starts punishing what cannot (yet) be explained.

The weird questions and the unexpected sideways connection, or the conversation with someone from another field that feels pointless for 40 minutes and then suddenly there’s a “DUDE THAT’S IT!” moment in the pub. That’s the serendipity of science and creativity that’s being phased out because “What if?” doesn’t get funded. Those fragile early ideas only sound stupid because all early ideas sound stupid until someone clever enough (or stubborn enough) gives them some love and attention. And systems obsessed with productivity are very good at crushing stupid-sounding, fragile things.

I’m sure they don’t always mean to, and plenty of pointless science trundles on. I don’t think there’s usually some villain in a lab coat standing over a stack of grant applications, cackling like Mr Burns, “Muwahaha, I shall destroy curiosity!” The system has come along over the years and we’re in a publish-or-perish period, so you need to look credible in front of people who have the money. And so researchers pick research topics that are more likely to get funded, which in turn means picking from a smaller pool of options and leaving the weird questions out on that weird limb. This is almost certainly due to the publishing model itself, but again, that’s another blog (you can read something better than anything I could put together over here, or listen to it over here).

“Being busy needs to be visible, and deep thinking is not. Academia has largely become a small-idea factory. Rewarded for publishing more frequently, we search for “minimum publishable units.” (2)

So, science is forced to justify creativity to people who can only recognise it once it’s already become useful because it’s been published and that is a big old ugly problem.

Creativity needs room, which is annoying because room costs money

You can’t just conjure up creativity at 10:30 am on a Tuesday because the department booked a creative ideation session with pastries and someone from comms has brought sticky notes.

“Right everyone, innovate.”

No, Piss off.

Creativity needs conditions, like time and space and autonomy and okay maybe some of those pastries. It needs people with different perspectives bumping into each other without already knowing what the answer is supposed to be. It needs the freedom to chase an odd thought without having to immediately explain its market potential in 300 words or fewer.

The paper makes this point, too:

Creativity depends on individuals and the environments in which those individuals work.

And that includes the physical and virtual spaces, the community around them, the autonomy they’re allowed and the chance encounters that let ideas jump across disciplines and labs. And that, I think, is called infrastructure.

And it doesn’t have to be buildings and machines and shiny equipment, that infrastructure might just be giving scientists (like all creatives) a little slack and some time to think. You need to be allowed to talk to the wrong people for the right reasons in a culture where saying “I don’t know yet” isn’t demmed a confession of weakness.

Science loves to pretend it’s above all this human stuff, but it isn’t, because science is done by humans, and humans are weird, social, anxious, brilliant, pattern-seeking mammals who occasionally need a walk, a chat, a biscuit and three hours of not being interrupted to do their best thinking. Like all creatives.

Science is not a bloody vending machine

Another problem is that science is often treated as something society deploys when it has a problem.

Pandemic? DEPLOY THE SCIENCE.

Climate crisis? DEPLOY THE SCIENCE.

Crime scene? DEPLOY THE SCIENCE.

Company needs a claim that sounds authoritative but doesn’t technically get them sued? DEPLOY THE SCIENCE.

Fine. Science can do that –it’s very good at solving problems. Scientists can make vaccines, analyse pathogens, build tests, model systems, assess evidence, improve policy, and tell Marketing departments that no, they cannot claim their moisturiser “rejuvenates cellular destiny”.

But science isn’t a vending machine for answers and you can’t just inserta problem, press button and receive certainty – that’s not how it works. Treating science like that makes everyone worse at understanding it. The public starts expecting clean answers, politicians start wanting certainty on demand, companies start using science as decorative authority and the scientists are trapped between the uncertainty of actual knowledge and the need to pretend there’s the certainty people seem to want/need from them.

I’ve previously written about scientists needing be more comfortable with communicating the inherent uncertainty of science, because science is temporal. It changes – as it should – based on new data and idea. I mean, that’s the whole point, it’s how we find out new things.

The paper points out that,

Though scientific work is logic and method driven, scientific knowledge is constantly evolving, dynamic and in a state of flux (1).

But that uncertainty and our ever-changing understanding is often ignored in favour of the comforting idea that science can always provide clear and definitive answers whenever someone important asks nicely.

Which is stupid. I mean, I get it, but it’s stupid. Because the real creative power is in noticing that the question is wrong. The creativity comes from reframing the problem and connecting fields that normally steer well clear of each other; it’s in seeing the system rather than the symptom. And that is much harder to sell.

“Give me the answer.”

“No.”

“What do you mean no?”

“I mean your question is bad and your thinking is worse.”

“Can we have the answer anyway?”

“Sure. It’s no.”

Expert generalists are not failed specialists

This bit from the paper made me want to punch the air and say “AHHH YEEAAAHHH”, or at least nod aggressively, which is the British middle-aged version. It talks about expert generalists. People who operate in between disciplines, who connect dots, identify patterns, improvise and move between fields rather than becoming the world’s leading authority on one tiny protein. These people are super valuable, but they’re difficult to fit into systems that reward specialist knowledge and clearly defined expertise.

AHHHH YEEAAHH– I mean, mmmm, yes, agreed.

Science needs specialists – of course it does. You want that person who knows every single wrinkle in that tiny protein’s weird little structure and what each and every amino acid is doing because they’ve spent decades looking at x-ray crystallography and sequencing data to the exclusion of everything else. But science also needs people who can wander between rooms and say, “This thing you’re doing over here looks suspiciously like that thing they’re doing over there, but with different acronyms.”

I know that sounds like I’m being flippant, but I kid you not, that’s a hugely difficult and creative skills and you want that person around. But our systems often look at those people and go, “Hmm, yes, interesting, but what are you exactly? What’s your speciality?” because if a thing can’t be easily labelled, measured, ranked, reviewed and placed neatly into a category, everyone gets nervous and just sweep under the Carpet of Inconvenience.

Science is art

Albert Einstein said that “The greatest scientists are artists as well,” and he may well have been right – bloke was right on a lot of stuff so I’m going to let him have that one. That doesn’t mean every scientist needs to start wearing a scarf indoors and describing their PCR as a meditation on temporal chaos. Please don’t, lab meetings are already long enough.

It means great science requires more than protocols and procedures. It needs imagination tempered by evidence, lateral thinking and vertical testing, an absolute shed-load of creativity; it requires you to look at messy data and not just ask “What does this show?” but “What else could this mean?” and “What am I missing?” and “How could I be wrong?” and “Why does this weird little result keep popping up?”; and it needs the space, the infrastructure, in which to do all of that.

Great scientists look beyond obvious answers and devise intelligent and truly novel experiments to test the nature of reality. Great scientists think about the breadth of possible explanations and interpretations from their data. Great scientists are deeply creative people because that is what great science DEMANDS.

Science is curiosity under pressure! It’s the messy, frustrating, beautiful act of trying to understand a world that always refuses to be simple.

So yes, science is creative. Obviously.

References

1. R. M. Morgan, R. L. Kneebone, N. D. Pyenson, S. B. Sholts, W. Houstoun, B. Butler, K. Chesters, Regaining creativity in science: insights from conversation. R. Soc. Open Sci. 10, 230134 (2023).

2. 1. D. Geman, S. Geman, Science in the age of selfies. Proc. Natl. Acad. Sci. 113, 9384–9387 (2016).