scientific theories possible—beyond the role of theories Some of science's most celebrated discoveries are at times seen as purely theoretical masterpieces of the mind, developed by brilliant thinkers without using tools and experiments. Take the classic example of quantum theory—describing the behaviour of subatomic particles at the smallest scales. At times, the theory is seen as uncovered by thought. But taking a closer look, the story shifts. Max Planck's quantum hypothesis in 1900, the starting point of quantum theory, did not emerge from just a flash of inspiration. He directly built on findings made possible by precision instruments like spectrometers and bolometers that measured emitted light across wavelengths with exceptional detail—namely the famous blackbody radiation experiments. And he applied the model of linear oscillators to explain energy exchange. These instrumental experiments did not just inspire a theory; it enabled and demanded one. The tools revealed the deficits that required a new interpretation. Einstein's work on the photoelectric effect in 1905 was not just a product of thought, but was firmly rooted in unexplained experimental findings using electroscopes and cathode ray tubes to observe how light liberated electrons from metal surfaces. He then interpreted what the tools had shown. Bohr's quantum model in 1913 was not just based on abstraction, but relied on rich data gathered with spectroscopes using prisms to reveal spectral lines—these pointed to electrons not orbiting randomly and exhibiting quantised energy levels. He also drew analogies between x-ray emission and electron transitions. Each of these scientists also relied on advanced mathematical methods. Even the unexpected discovery of the electron in 1897—made possible using the new cathode ray tube—provided further evidence of the need for a new atomic model. Using the newly invented interferometer, the Michelson-Morley experiment in 1887 indirectly reinforced the need to rethink light's behaviour. Far from being the triumph of just theoretical physicists, quantum theory only emerged from applying these powerful tools, along with others such as statistical mechanics developed since 1868. These remarkable instruments and real-world experiments uncovered what was previously invisible, enabling the theoretical explanation that energy and light might be quantised. Other iconic examples of what are at times celebrated as purely theoretical breakthroughs are Einstein'stheory of special relativity in 1905 and its extension into general relativity in 1915—describing objects affecting the curvature of spacetime. But these theories were also heavily grounded in unexplained experimental findings that key instruments revealed. In the 1887 paper On the relative motion of the Earth and the luminiferous ether, Michelson and Morley measured the motion of the Earth relative to the luminiferous aether that was thought to carry light waves. Surprisingly, the experiment demonstrated, by relying on the new interferometer, that the speed of light remained constant regardless of the Earth's motion through space. This extraordinary finding challenged the existing notions of absolute space and time. It laid the evidence that enabled developing Einstein's theoretical interpretation—that space and time could be relative. This landmark experiment was widely debated by leading physicists at the time, including Einstein who discussed the 'luminiferous ether' in his 1905 paper. And in Einstein's words: 'the Michelson and Morley experiment had made it clear that phenomena obey the principle of relativity' . Yet Einstein is said to have later downplayed the influence of these experimental findings on his thinking. Maxwell's equations in 1865 also laid the mathematical foundation for understanding electromagnetic waves, and led Einstein to reconsider how light behaved in different reference frames. Tensor calculus—a new mathematical method—was also indispensable for capturing the experimental findings and expressing gravitational fields in Einstein's equations. He did not invent these methods but adopted them. Discovering the electron with the cathode ray tube not only informed quantum theory but also highlighted inconsistencies in classical physics that relativity later resolved. Einstein's key role was in interpreting and restructuring these experimental findings that revealed what was not conceivable before the experimental findings. These extraordinary tools created the very possibility of seeing the world differently in the first place—and enabled Einstein to make the conceptual leap. Far from just confirming predictions, tools make new kinds of predictions possible. Technological advancements, including particle accelerators, telescopes, atomic clocks and gravitational wave detectors, then enabled precise experimental tests that validated and extended Einstein's predictions with remarkable precision. These again made new theoretical explanations possible that were not imagined before. Likewise today, theoretical physicists are fundamentally guided by experimental and methodological discoveries—for example by the particle physics experiments at CERN that inform their theories. Taking a step back, Einstein is famously credited with the most iconic equation in science that is associated with special relativity: E = mc 2. The simple formula states that the energy E of an object at rest (such as an atom) is equal to its mass m multiplied by the speed of light c squared. Similar mass-energy formulas, such as Е = kmc2, were already developed by Umov in 1873 and Preston in 1875, and later also by Thomson in 1881, Heaviside in 1889 and Poincaré in 1900. In 1904, Hasenöhrl derived E = 3/8mc2 while studying blackbody radiation—and his results were consistent with special relativity. In the 1903 paper published within the Proceedings of the Royal Veneto Institute of Science, De Pretto developed the very same formula E = mc 2, two years before Einstein's same version appeared in the German-based Annalen der Physik in 1905—the same journal where Hasenöhrl had published his work the year before. In 1906, Einstein then recognised Poincaré's mass-energy equivalence. And in 1907, Einstein acknowledged that others may have arrived at similar conclusions: 'It seems to me to be in the nature of things that other authors might have already elucidated part of what I am going to say. However, bearing in mind that the problems under consideration are being treated here from a new standpoint, I felt that I should be permitted to forgo a survey of the literature (which would have been very troublesome for me)' . Einstein critically connected the equation with the theory of relativity. But without Einstein we would still likely have a theory of relativity which other leading scientists, like Hendrik Lorentz, were also working on—and special relativity was originally called the Lorentz-Einstein theory. Darwin's theory of evolution from 1859 is often seen as a product of just reasoning and observation—as drawing sweeping conclusions from observing Galápagos finches and a long voyage. But in reality, Darwin relied on vast amounts of experimental data and scientific tools, both his own and those of others. His advanced microscopes—our prime example of a key scientific tool—enabled him to examine barnacles, corals and plant structures in great detail over the years. These shaped his understanding of variation, specialisation and adaptation across species. Darwin also conducted systematic experiments: he bred pigeons and plants under controlled conditions. These provided him with direct evidence of how artificial (human) selection could cause distinct traits to emerge across generations. These experiments did not just enable drawing the analogy for natural selection—they were also living laboratories for understanding how selection pressures shape species. The findings led Darwin to the conclusion: if humans could shape animals over time, so could nature. He also applied comparative anatomy methods to study finches, dissected animals and drew on the latest fossil discoveries—providing him with further evolutionary evidence of how extinct species relate to living species. And he was also influenced by Thomas Malthus' views on population growth—that we can compete for limited resources as a force of survival. By synthesising vast data across biology, geology and palaeontology and using natural experiments, Darwin was able to pull the pieces together into a coherent theory with unprecedented depth—one that transformed how we understand life and its origin. So why do the most celebrated scientific ideas and theoretical advances—quantum theory, relativity, evolution, DNA's double helix—commonly follow newly applied tools? It is because conceptual breakthroughs do not just deeply depend on advanced tools and unexplained experimental findings and puzzles, but are fundamentally made possible by them in the first place. Cutting-edge tools laid the essential foundation of evidence on which scientists could then create the theoretical explanation for what they could not have seen or conceived before. Here lies a key insight—tools do not just support the scientific process and make it more efficient, as commonly believed, they largely define what problems we can conceive and make our biggest scientific ideas and experiments possible. Without the microscope, no cell theory could have emerged; without x-ray crystallography, no DNA double helix model could have arose, and so on. Science is often seen and taught as theory coming first, with observations then merely testing them. But science's groundbreaking theories are commonly born from unexpected experimental observations. New tool → unexpected finding → new theoretical explanation of that finding. Theoretical physicists like Einstein claim that 'it is the theory which decides what we can observe' . Here, the broader evidence of discoveries across science leads to a deeper insight: it is the tools that define what we can observe, measure and experiment and ultimately what theories we can even conceive in the first place to explain surprising findings those tools uncover. So while in theoretical fields— like theoretical physics—observations often tend to be thought of as theory-laden, observations and theories are actually always tool-laden across science. Screening science's over 750 biggest discoveries, we find that only very few exceptional cases are popularly described as stemming from just 'thought experiments' . The idea that some theoretical breakthroughs can arise from just abstract thought is a somewhat romanticised view in the public imagination. The perception is especially common about theoretical physics, where scientists study at the smallest and largest scales (subatomic particles and the vast universe). That requires at times greater interpretation and abstraction from available evidence than is typical in most other fields across the natural and social sciences. In reality, such popular narratives of the lone genius are constructed only after new instrumental tools produced the unexpected findings and enabled interpreting those findings in a new light. Even Einstein's most famous and iconic description—imagining chasing a beam of light—was only part of his retrospective narrative decades later, not in earlier publications, interviews or communications. But we find that each of the few later descriptions actually fundamentally relied on pivotal enabling tools. The narrative came later; the tools came first—by revealing the puzzling problems that needed a theoretical explanation. Scientists—from Einstein to Darwin—are endowed with the same senses and motor skills as others, but they excelled at synthesising existing experimental findings and tools before developing their theory. And afterwards, major theories generally only become recognised as theoretical breakthroughs once they survive rigorous experimental testing with the tools that validate them—as we see scanning nobelprize discoveries. Even our best theories can only ever be confirmed using tools and methods. After all, Einstein received the Nobel prize in 1921 not for his relativity theory, but for his work on the photoelectric effect that did have clear experimental confirmation at the time. Next, when we explore how science's central tools emerged, we find that many were born largely as practical inventions using trial and error, experimenting and solving real-world constraints to our tools: from Kirchhoff and Bunsen's spectrometer, Martin and Synge's chromatography and Tiselius' electrophoresis, to Mullis' PCR method and Svedberg's centrifuge. From the first microscope and telescope born from lens crafting, to the thermometer based on the expansion of fluids; from Wilson's particle detector inspired by observations of cloud formation, to the Geiger counter built on observing ionised gases, a number of tools first began as hands-on applied breakthroughs rather than by building on an existing theoretical framework. For Röntgen's discovery of x-rays—an entirely unexpected observation—no theory was possible. Such tools were developed before theory even knew what questions to ask to develop them. But applying them then led to over a hundred major applied discoveries into invisible worlds, without yet theoretical explanations. So why has theory at times not been crucial for building some tools, especially first-generation tools? Because the goal is not explanation, but functionality—demanding testing and prototyping . But other central methods did rely in part on a theoretical foundation. Fisher's statistical methods solved real-world problems but also used practical principles of probability, and Ruska's electron microscope tackled real limits in resolution but also applied practical insights from quantum mechanics. When used to help optimise tools, applied theory is typically used—focused on solving real problems—rather than abstract theory, traditionally used to describe nature. Why are methods and tools so fundamentally important in propelling discoveries, often playing a more crucial role than scientific theories in science's over 750 major discoveries? Why are tools commonly the central driver in breakthroughs? There are several reasons: • Tools enable us to not only directly see, measure and experiment but also think in ways unimaginable without them. Remarkably, using them we can perceive previously unknown and inaccessible phenomena across all fields—from microorganisms and nerve cells to radio waves, subatomic particles and distant galaxies. By making the invisible visible, we have sparked hundreds of major discoveries through exploratory research and experimenting using new tools and with limited theoretical understanding. Tools—microscopes, x-ray devices, machine learning—observe and measure the unknown in ways that theories simply cannot, while we develop and revise theories (explanations of the world) from what tools reveal (Chapter 1). • Tools make discovery more strategic and less driven by chance. They shift discovery from chance to a deliberate and effective strategy—by enabling systematic, repeatable ways to explore unknown terrain. Uniquely, serendipitous discoveries, from x-rays to pulsars, generally reduce the role of theory as they are sparked by new tools of discovery before anyone knew how to explain them— turning observation into theory (Chapter 2). So tools often precede and inform the theories we develop. • Tools accelerate the pace and scale of discovery. Better methods—computational, experimental, statistical—generally evolve faster than the theories meant to explain what they uncover. New technologies like high-powered computing, machine learning and data analytics can speed up the pace of science. By collecting, analysing and visualising enormous volumes of data faster and with unprecedented precision, they allow us to test, refine and update discoveries including theories quicker and open new research areas (Chapter 1). • Tools push us past the boundaries of current theory, and without them our best theories could not have generally been imagined, much less tested . Conceptual breakthroughs occur after new tools are invented—from radio telescopes and particle accelerators to x-ray detectors. Because these tools unveiled invisible realms no one had conceived and enabled entirely new types of questions. Powerful new instruments reveal unexpected phenomena that current theories cannot explain. And theories do not explain themselves—they rely on the evidence that instruments provide. New evidence in turn forces us to reevaluate established theories. Detecting for example exoplanets using new spectrographs upended existing models of how planets form and sparked new theoretical explanations. And when it comes to the origin of our groundbreaking tools, we found that they generally emerge not from abstract theorising but from practical deficits in earlier tools and experimenting. • Tools spark breakthroughs across unrelated fields. Numerous transformative tools each triggered over a dozen discoveries—from electron microscopes and x-ray devices to spectrometers and chromatography—and often in unexpected fields far from where they emerged. What connects the distinct discoveries that each of these tools made possible is not a common theory, research teams, more funding or serendipity, it is using the common powerful tool—each unlocking groundbreaking findings across different fields. Versatile methods (like the gene-editing method CRISPR) are initially created in one field but then imported in others (with CRISPR transforming genetics, medicine, agriculture and environmental science). • Tools make science scalable, precise and reproducible. High-precision instruments enhance efficiency and allow researchers around the world to replicate experiments using standardised measurement. So the reason why tools often prove to be the more critical driver of discovery than theories is straightforward: the history of science shows they often not only resolve more of science's deepest questions but also spark more new fundamental questions and answers unimaginable before. Our tools foundationally shape not only what we know but how we know it. Across science, most fields are driven by experimental and observational methods and have a limited theoretical footing. Think of medical researchers and epidemiologists, ecologists and environmental scientists, archaeologists and palaeontologists, psychologists and the wide range of other applied scientists. When asked how discoveries are made, they highlight collecting data and running experiments using tools as the very cornerstone of science. A look across the fields of science shows that many disciplines thrive with little or no established theoretical foundation. Yet what unites all scientific fields is not theory—but methods and tools that ground knowledge in empirical reality. Without these tools, fields never get off the ground. No field can function without evidence derived from methods and tools. In some fields, researchers can have stronger synergies between technological advancements and the theoretical insights they spark, like with biotechnologists, chemists and materials scientists. At times, this synergy can help motivate tool design—with for example limitations in microscopy combined with quantum mechanics contributing to the scanning tunnelling microscope. In contrast, in a few subfields researchers place greater emphasis on theory, like theoretical physicists, theoretical economists and philosophers of science, often relying on mathematical methods with limited (direct) evidence. For these researchers, it is not surprising that the strongest emphasis on theory tends to come from fields where tools and experiments are used least as part of daily work. After all, they deal with phenomena far removed from everyday experience: like subatomic particles, cosmic scales and global financial markets—often expressed using mathematical equations. Yet even the most abstract theory must ultimately be based on and pass the test of reality. We require instruments before and after creating a scientific theory: before, tools provide the surprising observations and anomalies that spark theoretical ideas and explanations; and after, they offer the only means to test and refine them. A theory is always developed, directly or indirectly, by building on existing instrumental findings. No theory comes out of nowhere. And without the tools to create and verify theories with real-world evidence, they remain speculative—like string theory and the multiverse. Ultimately, both popular narratives—of a lone genius striking a fortunate flash of serendipitous insight or deducing fundamental truths through abstract thought— make science feel exciting and almost magical. By studying discoverers systematically, a different and more grounded story emerges: science's major discoveries are catalysed by new tools of observation and analysis and building on experimental work that enable the surprising insights. Surprise and insight can often be engineered. Here we provide a breakdown of the different types of discoveries and people's often simplified view of how science progresses. Ultimately, we find that most of science's over 750 major discoveries are experimental breakthroughs (60%), followed by method innovations (25%), and finally theoretical advances (15%). Theoretical advances awarded a Nobel prize are largely concentrated just in theoretical physics and theoretical economics—and are extremely rare, or non-existent, in fields like medicine and biology. Ultimately, without robust methods and tools to detect or measure, theories cannot generally be developed or tested but also more funding and collaboration are not enough, and serendipitous observations are not possible. Nothing can substitute tools in detecting, measuring and revealing surprising new discoveries. While the role of factors varies by discovery, tools are commonly the decisive factor—the common driver we must directly apply to be able to spark the breakthrough. Major advances often do not come from where we commonly look, but from where we do not commonly look—the methodological setup. Methods generate questions, data and evidence, and theories help shape the questions and interpret the data and evidence we generate. Theoretical discovery (classic view of scientific discovery) + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - relativity theory, quantum theory, theory of evolution methods, mainly mathematical, used by discoverer – and other new methods or tools used by other researchers – to develop a theory DNA's structure, hepatitis A virus, structure of insulin new method or tool used to reveal an experimental finding Experimental discovery electron microscope, statistical methods, particle accelerator new method or tool used to create a method innovation Method discovery x-rays, exoplanets, cells new method or tool used to make an unexpected finding Serendipitous discovery (romanticised view of scientific discovery) 14% 28% 58% 15% % of all nobel-prize discoveries only How discoveries emerge Common view of how discoveries arise Examples 100% % of science's major discoveries 15% 25% 60% 100% 16% findings are triggered by experiments, with or without a guiding theory (exploratory research) (common among: most scientists – from medical researchers and ecologists to behavioural scientists and other applied researchers) -tools are created by practical challenges and experimental work – with applied theory used, if any (common among: computer scientists, engineers, statisticians, bioinformaticians) findings are sparked by a flash of insight or accidental observation (common among: the public and some psychologists, historians and philosophers) theories are developed by abstract reasoning (but actually build on experiments and tools, and then require them for confirmation) (common among: theoretical physicists, economists, philosophers, the public) (usually an experimental/ observational breakthrough) Existing research – including tools, data, experiments and theories – is drawn on, to different degrees, in the discovery process Remarkably, many think methods and tools are stepping stones—serving their purpose until we create a discovery or theory, and are then forgotten. Yet we must learn to see our major methods as the core of scientific progress itself, and our discoveries and theories not as the final goal but as provisional outputs—providing provisional explanation and understanding. But they are always subject to refinement as we keep improving our methods that enable new evidence. The hope here is that we no longer just value science's major discoveries and theories as the pinnacle of science, but recognise science's great methods and tools of discovery as highly, if not more, as they make them possible in the first place.