Pioneering discoveries, such as the structure of DNA, the periodic table of elements and quantum theory, are embodied in our scientific fields, such as biology, chemistry and physics. A discovery captures an immediate advance, while a field captures our discoveries and methods—our ever-expanding body of knowledge over time. Scientific progress unfolds—and is measured—in three fundamental ways: through new discoveries, methods and fields. But how do new fields actually emerge? What is the key driver enabling them? Despite fields encompassing our immense scientific understanding, no comprehensive answer to this fundamental puzzle yet exists. Here we systematically trace the origins of science's major fields including hundreds of fields spanning across science, offering a unique opportunity to uncover what triggers new fields. Remarkably, we find that fields are kick-started by a new method or tool—from advanced telescopes to electrophoresis—as they enable a completely new perspective to the world and without them, the fields would not be possible. About a quarter of fields are the new method or tool themselves—from laser physics and computer science to x-ray crystallography and econometrics—forming entire disciplines around new techniques. Our extraordinary development of new statistical techniques, x-ray devices, microscopes and spectroscopes has fuelled a wave of new disciplines: each enabling over ten new fields to emerge. The common link uniting these diverse fields is not specific theories, large teams, more funding or even serendipity—it is that each relied on the same kind of powerful tool, used in remarkably different domains. The speed at which science expands is not random. The pace of opening new research domains is mainly determined by how quickly we create new tools: particle detectors launched high-energy physics, microscopy techniques triggered neuroscience and randomised controlled trials kick-started experimental economics. This simple yet powerful principle—if we begin to deliberately develop transformative methods—holds the key to accelerating scientific progress and enabling us to spark a tool revolution in science. The T
Scientific fields embody our vast, ever-expanding scientific, medical and technological advances and bodies of knowledge over time. But despite their tremendous impact, we still do not understand how new fields emerge. This is a fascinating and critical question because unravelling the mystery would enable us to explore entirely new domains. New fields can seem to arise from discoveries so groundbreaking that they create a new scientific community and body of knowledge. When we look closer, some new fields seem to emerge as a novel specialised branch of established disciplines, like molecular biology and developmental psychology. Others seem to grow out of the integration of two or more fields, like biophysics and cognitive science. Some seem to arise as contrasting approaches within the same domain, like theoretical and experimental physics and economics. Still others seem to develop due to shifting societal or environmental forces, like telecommunications and climate science. And these different paths can overlap. But what is the force enabling new fields in the first place, which then reflect greater specialisation, integration, diverging approaches and adapting to new external conditions? Given the diverse paths, can we identify a unifying explanation of how we give rise to all fields? If we could explain and even help predict their emergence, we would not just be passive observers of scientific growth—we could actively develop new research domains and speed up advances. The most common explanations for how fields emerge are through paradigm shifts in theories, new research programmes or splitting or merging scientific communities. But do they actually explain how new fields come about? A highly influential explanation, proposed by Thomas Kuhn, describes fields arising and evolving through fundamental changes in scientific theories—paradigm shifts. But many applied fields do not involve a theory. In fact, no new field emerged even from the classic paradigm shift from the Ptolemaic earth-centred theory to the Copernican sun-centred theory, or the shift from the theory of continental drift to the theory of plate tectonics to explain large-scale geologic changes. Another influential explanation proposes that major new research programmes lead to new fields. But many research communities—such as those studying new viruses, astronomical objects and global warming mitigation—have expanded our knowledge without spurring distinct new fields. Another influential explanation argues that scientific communities split or merge through new collaboration networks that can bring about new fields. But these shifts follow after a new field arises as an offshoot, rather than its driving force. Sociologists also offer explanations for how social movements and conditions can influence the scientific landscape and some disciplines. Yet, each of the proposed explanations is in fact mainly a consequence of new fields rather than their cause—they generally follow rather than precede their emergence. While researchers have long sought to explain new fields, efforts have mainly centred on research outputs, such as new theories and citation patterns, and aspects of fields like collaboration networks and productivity. But are 120 T HE ENGINE OF SCIENTIFIC DISCOVERY these really the most relevant metrics to focus on? In a thought-provoking article published in Science, researchers challenge this conventional approach: 'Citations, publication counts, career movements, scholarly prizes, and other generic measures are crude quantities at best … and their ability to predict the emergence of a new field or the possibility of a major discovery may be low' . Yet explaining and better predicting new fields is a key goal of the science of science. So if these standard metrics fall short in capturing the birth of new disciplines, what does capture them? Achieving this goal requires us to dig deeper and take a completely new approach. Here we shift our attention from commonly exploring outputs (especially article citations and theories) to investigating the inputs—especially new methods and tools that enable entirely new perspectives to studying the world. The idea that we can study the drivers of new fields to identify successful ways of advancing science, but that scientific fields have not yet been systematically analysed and linked to their underlying methods and tools, may seem ambitious. By tracing the origins of science's major fields including hundreds of different fields and the innovations in tools that enabled them, we show that this alternative approach is not only attainable—it may be the best strategy we have to understand how we develop new fields and how we can do so faster. We also explore the broader conditions behind the scientists kick-starting the fields. What do we find? A striking pattern emerges: about one hundred new methods and tools that won a Nobel prize have opened new fields, such as x-ray crystallography, electron microscopy, quantum computing, mass spectrometry, climate modelling, laser spectroscopy, phase-contrast microscopy and econometrics. Each of these is a powerful new method or tool, each earned a Nobel and each is the foundation of the new field it triggered—fundamentally reshaping the way we explore the world. Take the electron microscope, our prime example of a key instrument leveraged in multiple fields. Developed in 1933, this pioneering instrument does not just vastly magnify objects, it helped unlock an entirely new domain of biology: modern cell biology. By uncovering the intricate architecture of cells, it allows scientists to link cellular structure with function. It has transformed how we understand the causes of disease and enabled developing targeted therapies, paving the way for advances in regenerative medicine. The maser, invented in 1954 as the precursor to the laser, laid the foundation for new fields: laser spectroscopy and quantum electronics. It enables us to probe the dynamic behaviour of chemical reactions and the structure of molecules at an unprecedented level of precision. X-ray crystallography methods, created in 1913, gave birth to molecular biology. By unravelling the molecular structure of DNA and proteins, it reshaped our understanding of life itself—and remains a cornerstone for life-saving research in drug design and disease treatment. Exploring the birth of fields, we uncover a common pattern across them: whenever a scientific community undergoes a major change in the way it understands the world, it is preceded by a major methodological change in the way we study, measure and theorise about the world that was not possible before. T
To understand how fields emerge, we need to trace their origins and the pioneering scientists who sparked them. Here we analyse science's major fields including hundreds of fields spanning the physical, life and social sciences—from physics, chemistry and biology to psychology, economics, anthropology and computer science. Those less interested in the details of how we collected the data can skip this paragraph. We verified that all fields return at least 25 results on Google Scholar by searching 'field of x' (such as 'field of chemistry' or 'field of genetics'), with most generating hundreds or even thousands of results. To ensure all fields are recognised fields, we took several steps. First, we included all fields that emerged directly from nobel-prize-winning research in science—fields recognised in the prize motivation or opened by scientists who earned the prize. These make up 257 fields. Next, to broaden the scope, we also included all fields since 1500 pioneered by scientists featured in the seven science textbooks listing the world's 100 greatest scientists across science. These make up 116 other fields. This adds up to 373 fields and we then verified that the scientists who opened the fields are described in scientific publications as the 'founder' , 'father' or 'mother' of the discipline. Finally, we focus on fields established at universities, confirming that a total of 213 of the 373 fields—such as cognitive science, astrophysics and microbiology—exist as departments, institutes, centres or schools within universities. Each discipline is defined as a separate field or subfield—for example, the emergence of modern physics, then particle physics and later laser physics. (For emerging and smaller fields, additional analyses are available in the Online Appendix.) So what defines a scientific field? A field is commonly characterised by shared research topics, methods and tools, a scientific community and at times theoretical models. When a field emerged is established by the first article that kick-started its development—starting a growing research community and body of knowledge. The first publication that triggered each field is the main source for compiling the data; it includes the central method or tool applied in that publication, is commonly highlighted by the authors and made the field possible—and also includes the year they were published.
To begin, we examine the emergence of fields since 1500, tracing the year a new method or tool was created by the year the field arose applying it—unlocking an entirely new way to explore the world. Each groundbreaking new method or tool appears as a vertical line ( │) and each field that emerged using it appears as a dot (●)—seen in Figure 5.1. Think of for example the development of femtosecond (ultrafast) spectroscopy in 1985 (shown as a vertical line │) that led the Egyptian Ahmed Zewail to open the field of femtochemistry in 1988 (shown as a dot ●). This pioneering instrument enables us to much better understand chemical reactions on 122 T HE ENGINE OF SCIENTIFIC DISCOVERY very short timescales—mere quadrillionths of a second—using lasers. Suddenly, we could witness molecules in motion, reshaping our understanding of fundamental chemistry. Think of the invention of the microscope with the silver staining technique in 1873 that made it possible for the Italian Camillo Golgi and Spaniard Santiago Ramón y Cajal to launch the field of neuroscience that same year. This trailblazing technique revealed the first images of our nerve cells and enabled understanding the structure of our nervous system. It opened a new era in studying the intricate architecture of the brain and how we think, feel and remember. Think of the creation of a new 270-power microscope in 1673—capable of observing objects just one-millionth of a meter in length—that enabled the Dutch Antonie van Leeuwenhoek to give birth to the field of microbiology in 1674. For the first time in history, we could study the invisible world of bacteria and it even redefined what it means to be human. History reveals a key insight: new microscopy techniques—like new spectroscopes and controlled experimental methods—are representative of how new fields are born after inventing the right tools; and without them, they would not have been possible. We uncover a strong and direct link between new tools and the birth of new fields: examining the publications that kick-started science's new fields shows that each new field—one that studies and measures an unexplored part of the world—consistently begins by adopting a groundbreaking new method that enables studying and measuring the world in a new way. New method innovations precede the disciplines they trigger—with new meaning we use it for the first time ever for a specific problem. Some tools spark new fields almost instantly, while others lie dormant for decades before scientists pick up on the tool's potential. This relationship is mapped out here, where the length of each horizontal line reflects the time between the method's invention and the field it enabled (Figure 5.1). Two striking trends emerge: the lag between the two is shrinking, and the rise of new fields and bodies of knowledge is not slowing down (Appendix Figure 5.1). We uncover a very strong pattern here. The top ten methods and tools most used in nobel-prize discoveries have been so powerful that they each kick-started not one but five or more fields—in ways their inventors never anticipated: new statistical and mathematical methods gave rise to fields such as experimental economics and empirical finance. Spectrometers launched fields such as molecular spectroscopy and exoplanetary science. Electron microscopes triggered fields such as modern cell biology and electron microscopy. X-ray methods gave birth to fields such as molecular biology and x-ray crystallography. Chromatography enabled fields such as organocatalysis and signal transduction. Centrifuges kick-started fields such as enzymology and centrifugation. Electrophoresis enabled fields such as DNA sequencing and proteomics. Lasers sparked fields such as laser physics and laser spectroscopy. And particle accelerators and detectors paved the way for fields such as high-energy physics and solid-state physics. Beyond these top ten tools, others—such as geigermüller counters and game theory methods—have also given birth to at least five fields. These powerful, general-purpose tools explain most scientific progress because they each enable us to access a part of the world otherwise out of our scope. A powerful insight emerges from the fact that each of these tools kick-started multiple fields: what links these diverse fields is not a particular theory, research team, more funding or Methods and tools developed( ), and the fields they triggered ( ) 1500 1600 1700 1800 1900 2000 Year that main method/tool was developed 1500 1600 1700 1800 1900 2000 Year that field emerged using it Microscope, with silver staining technique, 1873 Neuroscience, 1873 Femtosecond spectroscope, 1985 Experimental economics, 1962 Randomised controlled trials/ Classroom experimentation, 1948 Microscope, improved high-power, 1673 Microbiology, 1674 Femtochemistry, 1988 Figure 5.1 New scientific fields emerge through new central methods and tools we create The data reflect 213 established fields developed since 1500 and the methods and tools that enabled them—with four examples illustrated. A new tool at times triggers multiple fields, appearing as multiple dots on the same horizontal line. We find that the year we create methods is closely correlated with the year the resulting fields emerge—94% of the variation. even serendipity—it is that each relied on the same powerful tool, used in remarkably different domains. What makes this connection astonishing is that none of these tools were specifically developed to open these fields, yet they still consistently sparked multiple fields across different disciplinary domains. Through quasi-experimental reasoning, we see that the new tools were not intentionally designed to create the new fields, but still did with their transformative power. The fact that entirely new domains of research emerged from tools designed for other purposes highlights a fundamental insight about scientific progress: new fields commonly come from the unexpected power of new tools. Without them, the new fields would not be possible. This unforeseen link between the two is key to understanding how science unfolds. Establishing that the new method (the independent variable) originally had no direct connection to the new field it ultimately enabled (the outcome) helps us isolate this causal link. And it reduces other explanations—such as funding, teams and institutions as the key triggers—as the necessary tools were already invented. So if we focus on outcomes in studying how fields emerge—like major new paradigm shifts in theories, research programmes, splitting or merging scientific communities, or even accumulated citations—we fall short in explaining the birth of disciplines. For these outcomes generally only arise at or after their emergence . But new tools, in contrast, can in fact explain and help predict new fields because they must arise before their emergence and be applied to kick-start them. This gives us a different way to think about this relationship: we need to track both the baseline (when tools are born) and the endline (when fields are born). Think of how randomised spectrography (1859)* microscopy with silver staining technique (1873) neuroscience (1873) x-ray analysis (1895) diagnostic radiography (1895) particle detector-cloud chamber (1911) high-energy physics/particle physics (~1900s/1910s) x-ray crystallography (1913) radiation genetics (1927) mass spectrometry (1919) particle accelerator/ cyclotron (1929) phase-contrast microscopy (1930) particle detector-Wilson cloud chamber (1932) electron microscopy (1933) econometrics (1933) molecular biology (1938) economic forecasting (1941) modern cell biology (1945) neutron diffraction/spectroscopy (1946) NMR spectroscopy (1946) protein crystallography (1946) particle detector-bubble chamber (1952) molecular spectroscopy (1957) photoelectron spectroscopy (1958) particle detector-hydrogen bubble chamber (1959) electron crystallography (1962) x-ray astronomy (1962) MRI (1973) cryo-electron microscopy (1981) scanning tunneling microscopy (1982) financial econometrics (1982) protein NMR spectroscopy (1985) femtosecond spectroscopy (1985) femtochemistry (1988) STED/super-resolution fluorescence microscopy (1994) frequency comb spectroscopy (2000) corporate finance (1954) single-molecule microscopy (2006) 1850 1900 1925 1950 1975 2000 Legend Scientific fields (those that are simultaneously a method or tool in italics) Method/tool that opened new field (but is not itself a field) Econometrics-driven fields Particle detector/accelerator-driven fields X-ray-driven fields Microscopy-driven fields Spectroscopy-driven fields microeconometrics (1974) Figure 5.2 New central methods and tools kick-start new scientific fields (illustrated with nobel-prize-winning methods) *The first spectrograph, developed in 1859, is the only tool included that did not receive a Nobel prize, but was used to open the field of mass spectrometry. T controlled trials designed in 1948 for clinical medicine foreshadowed the eventual (or inevitable) rise of experimental economics. This new branch of economics took shape when Vernon Smith adopted the transformative experimental method in economics in 1962 with his paperAn experimental study of competitive market behavior. With the creation of modern statistics and biostatistics in 1925, we could predict that the field of econometrics (statistical analysis in economics) would eventually develop—which took off in 1933 (Box 5.1). With the invention of the digital electronic computer in the 1940s and 1950s, we could predict that computer science would eventually arise—which did in the 1950s. The first particle detector that visualised particle tracks, built in 1911, gave birth to high-energy physics. The groundbreaking 1911 paper On a method of making visible the paths of ionising particles through a gas was published in the Proceedings of the Royal Society. The field then rapidly expanded with the creation of more advanced detectors, and also particle accelerators since 1929. These enable us to understand the nature of the particles that make up the very building blocks of matter and our physical world. Creating NMR spectroscopy in 1946 launched the fields of MRI and protein NMR spectroscopy. This helps us decode the structure and dynamics of proteins and develop life-saving medical drugs (see Figure 5.2). An improved cathode-ray oscillograph, invented in 1922, marked the birth of neurophysiology, helping unravel how our nervous system functions and diagnose and treat neurological diseases. Game theory methods, first conceived in 1928 and expanded in 1950 with Nash equilibrium, reshaped economics by giving rise to information economics—a field that explains how information impacts our decisions and economy. The transistor, a tiny semiconductor device built in 1947, made the field of microelectronics possible, bringing about mobile phones and personal computers and laying a key foundation of modern society. It was created at Bell Labs by John Bardeen and Walter Brattain and later refined by William Shockley. At the time, they had no idea that the transistor would forever change the world and kick-start what has become a massive field that shapes our everyday lives through the technologies we use. They set off a technological transformation that has changed the way we communicate, work and live. Each of these groundbreaking tools earned a Nobel prize. Without innovating new methods, these new research domains—and the world-changing technologies they enabled—would not have been possible. Box 5.1 How Ronald Fisher developed modern statistics—enabling powerful analysis across science and opening new fields We now zoom in on statistical methods—one of the ten central methods and tools most used to make nobel-prize discoveries and trigger new fields—before zooming back out to the broad patterns across science. Statistics transforms how we study the world by allowing us to collect and analyse vast amounts of data and run large experiments on almost any phenomenon—from tracking diseases in populations and decoding cellular processes to mapping astronomical bodies, 126 T HE ENGINE OF SCIENTIFIC DISCOVERY modelling complex economies and even uncovering what shapes discoveries and fields. The foundation of modern statistics developed in stages around the early 20th century, culminating in 1925 with the now classic book Statistical methods for research workers. This groundbreaking book, written by the London-born Ronald Fisher at the age of 35, marked the first full-length book on statistical methods, playing a pivotal role in establishing and spreading modern statistics. Fisher's path to transforming statistics was anything but conventional. After earning a bachelor's degree in astronomy and teaching several years in public schools, he received an offer in 1919 to work with the renowned statistician Karl Pearson. But due to a long-standing feud between the two, Fisher declined and instead joined Rothamsted, an agricultural research institution. Here he ran plant-breeding experiments using and refining statistical techniques. This hands-on research environment provided the inspiration for his 1925 book that offered solutions to many real-world problems his colleagues faced with handling data. In his book, Fisher writes: 'Daily contact with statistical problems as they presented themselves to laboratory workers stimulated the purely mathematical researches upon which the new methods were based' . And after Peason retired in 1933, Fisher took over his position as chair at University College London. In this landmark book, Fisher developed some of today's most widely used statistical methods across nearly all fields. Among his most influential contributions, he created the analysis of variance (ANOVA) technique that enables us to assess the average differences among sub-groups or experimental conditions within a study. Whether comparing the effects of different medical treatments, agricultural practices or economic policies, ANOVA became an essential tool for detecting meaningful patterns in data. Fisher also redefined how we determine statistical significance by popularising the p-value, or probability value. He proposed using a 1 in 20 probability threshold as a standard to evaluate whether a result arose by chance in an experiment. He wrote: 'The value for which P = .05, or 1 in 20 [chance] … it is convenient to take this point as a limit in judging whether a deviation is to be considered significant or not' . This threshold remains widely used in scientific research, though debates continue over whether 0.05 is an optimal cutoff (Picture 5.1 ). Fisher also pioneered using the important technique of randomisation in experiments—a technique to reduce bias by selecting samples randomly rather than arbitrarily. He applied the principle first in his agricultural experiments. Remarkably, he developed these pioneering methods through his everyday work and using little more than a calculator, without a team or additional resources. His statistical breakthroughs transformed fields like agriculture and biology and paved the way for others, like medical statistics, econometrics and complexity science—demonstrating how powerful new methods can ripple across disciplines in ways its creator never anticipated. He also played an essential role in founding population genetics and shaping the modern evolutionary synthesis—merging Darwin and Wallace's theory of evolution with Mendel's heredity. Fisher is seen as the father of modern statistics and even as 'the greatest of Darwin's successors' . T F or his contributions, he was knighted in 1952. Yet he remains one of the greatest unsung heroes of science: most scientists have likely never heard of Fisher, even though nearly all scientists have applied his methods. His work also set the stage for randomised controlled trials, a method pioneered by Bradford Hill. The method is one of the most important statistical innovations in fields like public health, economics and psychology—essential for testing the effectiveness of medicine we take and policies we adopt. Fisher did not just refine statistics, he redefined how we think about and understand the world. p = 0.05 z = 1.96 –3 –2 –1 0 13 Picture 5.1