Introduction — Sparking a method revolution in science

Our powerful new methods and tools trigger breakthroughs by enabling us to see, measure and imagine the world in ways that were impossible before them. Microscopes and telescopes opened hidden worlds too small or too distant for the naked eye—revealing unexpected parasites, bacteria and nanoparticles, and uncovering new distant star clusters and galaxies. Statistical methods stretch the limits of our mind, allowing us to process and analyse enormous amounts of data. They uncover new patterns we could not see otherwise—from tracking gene mutations to identifying climate trends. X-ray, MRI, ultrasound and CT devices make it possible to generate extraordinary images inside our body without a single incision. They have revolutionised medicine by letting us better diagnose, monitor and understand disease. Spectrometers and radar devices enable capturing and measuring light, radio waves and signals that our senses cannot detect—from astronomy to cancer detection. Controlled experimental methods allow us to test treatments for countless 6 T HE ENGINE OF SCIENTIFIC DISCOVERY diseases—from tuberculosis to scurvy—in rigorous ways that have saved millions of lives and were unthinkable before. Computer technology lets us simulate complex systems—from the earth's climate to neural networks—and even predict the spread of pandemics that would otherwise be too vast to grasp. AI and machine learning methods enable analysing immense datasets, can perform experiments and make predictions based on spotted patterns—with exceptional speed. New tools give us new questions to ask and new ways to answer them. Each powerful method reveals a new layer of reality that we often did not even know existed. In fact, the efforts of millions of minds, eyes and hands studying the world are rarely as powerful as someone developing or getting their hands on a new tool. A single invention—whether an advanced telescope or x-ray method—gives us a new lens to perceive and probe further and deeper than humans can alone. It gives us a new way to trigger multiple discoveries with a single tool. We use the microscope, throughout the book, as a prime example of a powerful discovery tool—transforming what we are able to uncover. The first microscopes radically amplified our limited vision to reveal a vast microscopic world we did not expect. This optical breakthrough was continually extended and refined over time. Eventually came the electron microscope: one of the most powerful instruments in the history of science. A 26-year-old German graduate student Ernst Ruska invented the instrument in 1933 at the Technical University of Berlin. By using beams of electrons instead of light, the electron microscope depicts images of miniscule objects. And it vastly expanded our world of microorganisms, molecules and nanoparticles. In fact, this tool sparked over a dozen major discoveries—from the cell structure in 1945 to ribosomes in 1955. What unites these discoveries is not a shared theory, research teams or more funding, it is a shared tool—the electron microscope—that made these breakthrough findings possible across different fields. It also launched new fields like modern cell biology that would have been impossible without it. Here we uncover how the story of science is largely driven by inventing and reinventing powerful methods and tools—from CRISPR gene editing to space telescopes—that trigger our big breakthroughs. What are the top ten methods and tools most used to spark discoveries? Examining science's over 750 major discoveries, we identify the top ten—and highlight them in Figure 0.1. Interestingly, the top ten methods and tools are almost the same when looking at all nobel-prize discoveries only, except that particle accelerators and particle detectors make it into the top ten, while telescopes and thermometers do not. Take electrophoresis and chromatography methods—before inventing them, we faced many barriers to analysing chemical substances. But once developed, they enable us to separate and determine the composition of DNA, protein molecules and viruses— in ways impossible before. Electrophoresis was developed by the Swedish graduate student Arne Tiselius at Uppsala University—at just 28. It enabled making major discoveries and another powerful method: DNA sequencing. Then there is partition chromatography—one of the most important methods in chemistry and biology. It was invented by the British graduate student Richard Synge at just 27, alongside his colleague Archer Martin at 31. By separating and analysing complex chemical mixtures, this extraordinary method enabled discoveries from the structure of insulin to how carbon dioxide is absorbed in plants. Today, we rely on it IN in many areas that affect our everyday lives—from pharmaceuticals to food safety and environmental testing. Surprisingly, Archer Martin, in his Nobel prize speech describing the method, said that: ' All of the ideas are simple and had peoples' minds been directed that way the method would have flourished perhaps a century earlier' . What is remarkable is that we find many of these cases across science: the worldchanging potential of tools that can sit unnoticed for decades simply because not enough researchers are dedicated yet to developing tools. Yet adopting tools much later comes with important opportunity costs—and delayed impact for science and society. It is also striking that these groundbreaking inventions—among the most important tools in science—were made by young graduate students doing handson experimental work as part of their PhDs. This tells us something crucial about how discovery works: progress is often not about greater funding, bigger teams or big theories. But about researchers refusing to accept the limits of existing methods and tools, and building new ones that unlock entirely new ways of seeing, testing and understanding—as we will explore throughout the book. E=MC2 (electron) microscopes x-ray methods lasers statistical/mathematical methods spectrometers thermometers chroma

tography centrifuges telescopes electrophoresis New methods and tools New discoveries Figure 0.1 New methods and tools trigger new discoveries Here we will be unpacking the fascinating stories of how and who developed our most powerful tools of discovery—and how they have revolutionised science by enabling completely new lenses to the world. Our ability to understand the world is limited by the tools we have developed so far. By the ways we can presently observe, measure and uncover—whether neural synapses, viruses, radio waves, the human genome or exoplanets that were unimaginable before our tools. For the tools we have created up to now shape the questions we can ask, the answers we can provide and the way we can explore and understand the world. 8 T HE ENGINE OF SCIENTIFIC DISCOVERY But are there major theoretical breakthroughs that are mostly just sparked by intuition or insight? What about some great theories of science—are relativity, quantum mechanics, evolution mainly a product of thought? Some believe they are. Here we will see how Einstein developed the theory of relativity by building on unexpected experimental findings, made possible by an innovative new instrument: the interferometer. Those puzzling findings showed that the speed of light remained constant regardless of the Earth's motion through space. This extraordinary evidence challenged the existing notions of absolute space and time. It enabled developing Einstein's theoretical interpretation. In Einstein's words: 'the Michelson and Morley experiment had made it clear that phenomena obey the principle of relativity' . Tensor calculus, a new mathematical method developed at the time, was also indispensable to capture the experimental findings and express gravitational fields in Einstein's equations—what we will unpack later. To develop quantum theory, Max Planck also did not just rely on thought, but took the first step by analysing unexplained findings from experiments on blackbody radiation, made possible with powerful tools—spectrometers and bolometers—that measured emitted light. Einstein's work on the photoelectric effect then contributed further and was rooted in experiments using electroscopes and cathode ray tubes that showed how light liberated electrons from metal. Then came Niels Bohr's quantum model that relied on spectroscopes using prisms to reveal spectral lines—that pointed to quantised energy levels. To create the theory of evolution , Darwin also relied on tools like his advanced microscopes that enabled him to study barnacles, corals and plant structures that shaped his understanding of variation and adaptation. He also conducted breeding experiments with plants and pigeons, providing him direct evidence of how artificial (human) selection could drive changes in species over generations. These experiments enabled him to draw the analogy for natural selection. Far from being an abstract creation, science's greatest theories—from relativity to quantum mechanics to evolution—are only possible by relying heavily on key unexplained experimental findings that scientific tools revealed. These extraordinary tools created the possibility of seeing the world differently in the first place. They laid the essential foundation of evidence that made it possible to develop theoretical explanations for what they observed. Far from just confirming theories, tools make new kinds of theories possible. Yet in fields like theoretical physics, some exceptional theoretical advances have involved larger conceptual leaps—leaps from existing experimental findings and tools to theory—than in most other fields across science. We will trace the origins of science's biggest breakthroughs in history as we work through the book. Taking a step back, a striking insight emerges: each method and tool gives us a unique lens to seeing the world. The sum of these lenses largely determines how far we can currently see, explore and think—our present limits of science. To better understand the world and science, we have to better understand the tools we invent to study the world and do science. Because accelerating discovery depends on how well we understand our powerful methods and tools and how we develop them—as they lay the foundation that science is built on. So how do we build strong evidence and develop a general theory? The National Research Council in the US confirms: 'No theory of scientific progress exists … that I allows prediction of the future development of new scientific ideas' but researchers instead often emphasise how uncertain discovery can be. But there is a better approach to tackle the puzzle—one grounded in rich evidence. In general, there are two powerful ways to develop strong evidence and a general theory. The first is to explore the entire landscape—the total population as comprehensive as possible, not just a subset. That is what we do here: rather than typically studying a sample, we examine science's over 750 major discoveries across fields—all nobel-prize and major non-nobel discoveries. (We will unpack these discoveries in the next chapter.) By studying this much larger population, we get a clearer, more accurate picture of what actually drives discovery—by spotting the general patterns. In other words, what science's biggest discoveries have in common—and what they do not. And these two independent data sources enable us to test and compare the strength of the patterns and findings. The second way is to approach the problem from multiple perspectives. Instead of relying on a single method, we use different methods and perspectives from diverse fields—from statistics and history of science to cognitive science, and beyond. This gives us multiple, independent ways to test and confirm the patterns we find—and the strength of the findings. Rather than just relying on one method that can be more narrow or have blind spots. This combination—using the most comprehensive data possible and diverse methods—lets us fill the gap: a general theory of how discovery emerges. Here we develop the new methods-driven discovery theory. What the theory explains is straightforward: new methods and tools unlock new discoveries by enabling us to see, measure and understand the world in ways that are impossible without them. The ultimate test of a successful theory is: can it explain what drives science and discoveries—and can it advance and better predict discoveries? On this test, this theory can take us further than current explanations—as we will see. It can explain our past discoveries—and improves predicting future discoveries: the speed and direction of method innovations we make give us powerful signals that breakthroughs are soon to follow and where they can come from. The central thread connecting the different chapters of the book is exactly this expanding toolbox of ours.