Scientific reasoning requires irrationality of intuition

SUMMER SERIE: HOW SCIENCE WORKS

Science is based on coherent reasoning, while intuition follows a more disjointed path. But scientific research couldn't succeed if scientists didn't listen to their gut feelings every once in a while.

Science by definition relies on logic, reasoning and rigor. But we've all heard about Archimedes' famous Eureka! moment when, while taking a bath, he suddenly understood what came to be known as Archimedes' principle. While this anecdote may or may not be true, it does illustrate the catalyzing role that intuition can play in scientific research.

Intuition isn't logical and doesn't always manifest itself clearly. It's that light bulb moment when the data suddenly make sense.

At the EPFL lab run by Bruno Lemaitre, an immunology professor and head of EPFL's Global Health Institute, scientists study the genes of Drosophila flies in order to better understand their immune responses and gain insight into the immune systems of mammals. The scientists are very methodical in their work, yet they still make room for intuition. "Our research is similar to a police investigation," says Lemaitre. "We test out various hypotheses, and if they prove to be false, we use our imagination to come up with other ideas. Then our intuition helps us sort through the various ideas and see if any of them could explain what we've observed. Finally, we perform analyses to check whether our new hypotheses make sense."

Research decisions shaped by intuition

In the end, it's the scientists themselves, and their complex human nature, who keep science alive and push it forward. "Research - like any other activity - is based on impressions and feelings," says Lemaitre. "If all we did was execute logical tasks like a computer, we wouldn't make any new discoveries."

We use intuition to discover and logic to prove.

So how exactly does intuition factor into the research process? Frédéric Blanc, an EPFL professor and particle physicist, believes it guides many of scientists' decisions. "For instance, if we want to build a particle detector, we'll read the literature and draw on other physicists' experience to help us choose what kind," he says. "But we'll rely on our intuition to select which type of model to test first. And when it comes time to select the detector's size and geometry, our intuition will help us make the inevitable trade-offs, deciding which parameters to prioritize over others."

Blanc is currently studying particle interaction as part of CERN's Large Hadron Collider beauty (LHCb) experiment, which generates dozens of terabytes of data. "As researchers, we have to have a feel for which analysis method would work best for our data," he says. Fingers crossed that the team at CERN has chosen the right one, because the kind of B-meson decay they're interested in occurs only once out of a billion times.

Intuition also influences the direction in which scientists choose to take their research. That became apparent in a study conducted by EPFL's Laboratory of Molecular Simulation (LSMO). This lab is specialized in metal-organic frameworks (MOFs), which are porous chemical "sponges" made from metal ions and organic compounds and used to absorb certain types of gases. However, synthesizing and simulating new MOFs is an arduous process that could require millions of experiments to create just one MOF. Because scientists don't have that kind of time, they have to make certain choices upfront as to which experiments to run. To explore how scientists make those decisions, the LSMO team used a high-performance robotic synthesizer to reverse-engineer the process of developing the MOF called HKUST-1. The robot compiled all possible failed or partially successful experiments for creating an HKUST-1 molecule. The team found that even if the robot processed around 30 reactions each day, it would still take nearly 27 million years to run through all possible reaction combinations. This shows the benefit scientists gain from using their intuition, and that of their colleagues, to orient their research.

Collective intuition

According to Lemaitre, scientific discovery often results from combining the intuition of several experts. "When you have a group of people working together in a lab who are all enthusiastic about a given topic, there's a good chance something will come out of it," he says. That's because they'll be constantly talking about it, and these discussions will give rise to new ideas. "In this case, a new discovery could be the culmination of the group's collective intelligence," says Lemaitre. His colleague Blanc also points to the water-cooler effect. That is, "the casual discussions where people chat spontaneously about their work, without necessarily expecting it to lead anywhere," he says. "That's typically when breakthroughs happen. It's a sort of collective intuitive development."

Teaching and nurturing intuition

Fortunately, intuition isn't just a gift you're born with. At EPFL, it's even taught. Donna Testerman, a researcher and professor in the mathematics section, tells her students that it's important to cultivate their intuition. "Many students think that math is a rigid subject that entails merely repeating what others discovered before them. What they don't realize is new discoveries are happening in math all the time, in the same way as in any other field," she says. "If you think you may be onto something, you naturally want to prove it. And by attempting to prove it, you might come across an obstacle, which leads you to find counter examples. And by the end of the process, you've figured something new out."

By the same token, Blanc encourages his students to stay humble about their intuition. "At first students will say things like, 'I think this is how it works,' but then I point out that what they're saying is just an assumption. If they're right, great. If not, they need to ask themselves why they were wrong." Intuition can thus also be the starting point for thinking more broadly about a problem. And developing your intuition means feeling your way, step by step, to the solution.

Students need to be humble about their intuition. If they're wrong about something, they should ask themselves why, get back on the right path, and keep practicing.

There's another approach Blanc also takes with his students - a fairly pragmatic one for a professor in an abstract subject like particle physics. He encourages them to visualize the concepts they're studying. "For instance, if you try to imagine a particle detector and the particles traveling through it - the curve they follow, their trajectory, etc. - and you can visualize all that in your mind, that will help you grasp the physics behind your experiment, better understand its results, and, above all, guide your intuition."

Lemaitre often finds himself going off on a tangent during his lectures, telling his students about what life is like as a scientist and what goes on in a research lab. "Some students love it, while others ask me to stick to the topic," he says. He believes that it's another way to nurture students' intuition, which is built through experience, the knowledge they acquire and how they view the world around them

Do Eureka moments really occur, or is it just marketing?

Lemaitre and Blanc both agree that there are indeed moments when the light bulb goes on and you suddenly understand something, without really knowing why. "It's like you've jumped across a logical divide," says Lemaitre. "And there's no going back. Everything takes on a different meaning, and once you've understood how a system works, everything falls into place

So intuition is clearly one of the core elements of scientific research. That said, the process is a little more complicated than having an apple fall on your head and suddenly understanding gravity.

/Public Release. This material from the originating organization/author(s) might be of the point-in-time nature, and edited for clarity, style and length. Mirage.News does not take institutional positions or sides, and all views, positions, and conclusions expressed herein are solely those of the author(s).View in full here.