The Structure of Scientific Revolutions - critical summary review

When it was first published in 1962, “The Structure of Scientific Revolutions” by American physicist and philosopher Thomas Kuhn caused a strong stir in the philosophy of science community. By offering one of the first history-aware and theoretically grounded explanations of scientific change, the book challenged the heroic view of scientific progress as an incremental process, while articulating a highly developed alternative view. Still controversial, this view has been extraordinarily influential, both within philosophy and outside it.

According to Kuhn, rather than being uniform and accumulative, the development of science is actually episodic, alternating between long periods of puzzle-solving “normal” science and short bursts of anomaly-driven, paradigm-shifting “revolutionary” research. Get ready for a 12-minute clarification of the previous sentence and prepare to hear an interesting theory of how and why scientific revolutions take place.

Scientific progress vs. history

Science, as defined by Encyclopedia Britannica, is a “system of knowledge that is concerned with the physical world and its phenomena and that entails unbiased observations and systematic experimentation.” It is, inarguably, humanity’s most organized attempt to discover the general truths and the fundamental laws of the world we live in. The scientific method involves careful observations, measurement-based testing of deductions and hypotheses, and the modification or rejection of the hypotheses based on the experimental findings. 

Combine all of this with rigorous skepticism, and it’s only natural to assume that science progresses gradually and incrementally. In other words, new generations of scientists build upon the findings of their predecessors, and so on. Perhaps nothing captures this idea better than Isaac Newton’s 1675 letter to Robert Hooke, wherein the great English polymath states, “If I have seen further, it is by standing on the shoulders of Giants." Unsurprisingly, this is the current motto of Google Scholar, perhaps the most widely used online bibliographic scholarly database.

But there’s a problem with this view, argues Kuhn, and the problem is historical. Put simply, history doesn’t support the idea of incremental scientific progress. Quite the opposite – the more carefully one studies the history of science, the less certain he or she becomes that science develops by the accumulation of individual discoveries and inventions. For example, Copernicus’ heliocentric model of the universe didn’t utilize anything of the Ptolemaic geocentric system. Instead, it contested the idea of the Earth being the center of the universe. Just as well, Aristotelian mechanics was made obsolete by Newton’s discovery of the classical laws of motion, which, in turn, were disproved at microscopic scales by quantum physics. To stay with Newton, during the 18th and the 19th centuries, his law of universal gravitation was considered the definitive scientific explanation of gravity, but today it has been superseded by Albert Einstein’s theory of general relativity.

Now, if both Newton and Einstein were scientists, doesn’t that imply that incorrect scientific theories can be produced “by the same sorts of methods and held for the same sorts of reasons that now lead to scientific knowledge”? Kuhn claims it does, because, rather than producing definitive scientific explanations, the scientific method is designed to produce approximations – that is, frameworks that help scientists organize further research. Kuhn calls these frameworks “paradigms” and defines them as “universally recognized scientific achievements that, for a time, provide model problems and solutions to a community of practitioners.” To understand paradigms better, it’s essential that you understand the four key phases of scientific development.

Phase 1: pre-paradigm science as chaos

Science starts with speculation. However, as long as there is speculation, there is no science. There is only “fact-collecting” and “frequent and deep debates over legitimate methods, problems, and standards of solution.” Kuhn calls this period of fact collecting and debates that precedes scientific consensus and objective-oriented research the “pre-paradigm” phase of scientific revolutions.

During the pre-paradigm phase, intellectual activity is fairly disorganized and diverse. There are as many theories as there are theorists, and competing schools of thought use different theory-dependent procedures to “prove” their observations, leaving no room for collective progress. For example, before the scientific revolution of the 19th century, there were many scientists debating phenomena such as heat, magnetism, and electricity, and each of them believed they were right. However, in the absence of a commonly accepted observational basis, it was impossible to weed out the irrelevant observations from the important ones – that is, there was no way to sort things out. 

It may be difficult to understand the pre-paradigm phase today when there are so many well-formulated theories in so many different areas, but just a quick look back at Ancient Athens may be enough to give you a proper idea. With philosophers lacking a clear understanding of how the world works, in the span of just a single century, they thought up numerous different and conflicting ideas about the origin and structure of the cosmos, and almost each of them produced a different school of thought. Was fire or water arche, the first principle of all things? Or maybe it was air, or some other unknown substance of unlimited extent and duration? What about the so-called four elements working together? 

These questions may seem strange today, but just consider the current state of quantum physics. There are literally hundreds of competing quantum models and theories, and we are yet to witness a groundbreaking experiment or hear of a breakthrough observation that will make some of them obsolete, and draw the majority of researchers to one particular problem-solution. Throughout history, it was breakthroughs of this kind that transformed speculation into scientific research, and made it possible for “normal science” to appear.

Phase 2: normal science as within-paradigm puzzle-solving

Once the majority of researchers are drawn to a proposed solution to a particular problem, a widespread consensus is formed around the relevant theory. The consensus allows scientists to finally agree on fundamentals – such as scientific procedures, instruments, language, and theories – which, in turn, allows them to move from chaotic fact collecting and purposeless debates to purposeful research. It is the consensus, the agreed-upon conceptual framework, that gives scientific activity purpose, by defining what is possible and rational for a scientist to do. For example, before Copernicus, it was only sound for a scientist or natural philosopher to assume the Earth was the center of the known universe, and try to solve other problems without challenging this fact. However, today it would be irrational to base your research on a similar idea.

Kuhn calls the consensus (that is, the conceptual framework that makes science possible) a paradigm. When there are many paradigms floating around – as in Ancient Greece or today’s quantum theory – we are talking about pre-paradigm speculation. However, when one of these paradigms proves to be simpler or more complete than the others, and thereby becomes dominant, the second phase of scientific progress commences. Kuhn calls this phase “normal science” and describes it as “the activity in which most scientists inevitably spend almost all their time.” 

Normal science is stable, but it is also kind of predictable. Rather than trying to discover something new, most scientists operate under the assumption that they already know what the world is like, and that their job is to merely prove the paradigm they have inherited. That is why they spend most of their time either articulating the dominant theory via mathematical analysis or tidying it up where known observations don’t quite match with it. Rather than inviting defects – as Karl Popper’s theory of falsifiability suggests – normal scientists merely try to eradicate them; and rather than being impartial, they are pretty biased toward the dominant paradigm. After all, they want to prove it, not disprove it. In fact, “the most striking feature” of normal science is one that goes against the common idea of the scientist as an innovator. Namely, normal science doesn’t aim to produce major novelties, but to solve paradigm-related puzzles.

Phase 3: anomalies and scientific crises

“Under normal conditions,” explains Kuhn, “the research scientist is not an innovator but a solver of puzzles, and the puzzles upon which he concentrates are just those which he believes can be both stated and solved within the existing scientific tradition.” And indeed, almost all scientists do only experiments, the results of which they can guess in advance. That’s why most experiments start with a hypothesis, that is to say, a conjecture of the possible outcome. Even more strangely, most experimenters try to prove their hypothesis and are disappointed when they cannot. In Kuhn’s explanation: “The man who is striving to solve a problem defined by existing knowledge and technique is not just looking around. He knows what he wants to achieve, and he designs his instruments and directs his thoughts accordingly.”

However, as much as normal science might try to clear up the status quo within an inherited paradigm, one day, a certain experiment will yield an unexpected result – an observation that cannot be explained within the given conceptual framework. To account for the anomaly, a scientist may try to extend the framework. For example, even though Ptolemy’s geocentric theory couldn’t explain the evident “wandering” of the planets in their orbits, this age-old observation was forcefully made to fit the dominant scientific paradigm through the addition of smaller rotating spheres upon the major ones, called “epicycles.” For reasons ancient scientists couldn’t have even guessed, epicycles worked very well and were highly accurate in predicting the speed and direction of the moon, sun and planets. Thanks to the epicycles, the anomaly remained an anomaly, and the paradigm didn’t change.

Here’s another more recent example of an anomaly. A few years back, scientists at the CERN research center in Geneva – which was, more or less, established to experimentally prove Einstein’s theory of relativity – announced that they had managed to observe faster-than-light neutrinos. In the existing Einsteinian scientific paradigm, which has helped us organize our existing knowledge pretty well, there is no such thing as speeds faster than light. So, the observation had to be either an anomalous one or one that proved Einstein wrong. A brief scientific crisis followed, during which the majority of scientists vehemently claimed that Einstein couldn’t have been wrong, and that it was the experiment that was flawed. In the end, it turned out that they were right and, consequently, the current scientific paradigm survived. However, Ptolemy’s geocentric model and its epicyclic extension could only survive until Copernicus, who provided the basis for a better and simpler explanation of planetary motions.

Phase 4: scientific revolutions and incommensurability

Whenever they appear, observational anomalies question existing scientific paradigms. Even so, they are usually tolerated for a while, as scientists believe they will be able to explain them over time. They have all the right to, because this has often been the case throughout history. However, it sometimes happens that an anomaly cannot be resolved within the existing paradigm, no matter how much that paradigm is modified or adapted. In cases such as this, a scientific revolution happens, or – as Thomas Kuhn calls it – a paradigm shift. 

A paradigm shift is a major, radical and fundamental change in the core concepts and practices of a given discipline or domain. As a result, it produces not only new kinds of puzzles for forthcoming generations of scientists, but it also generates a completely new way of looking at the world for the majority of people. For example, more than two millennia ago, Aristotle observed that heavier objects fell faster than lighter ones, and that some objects didn’t even fall. To explain this observation, he surmised that different objects were made of different elements (earth, air, fire, and water), and that these elements had a propensity to move back toward their “natural place.” Thus, if an object was made mostly of earth, it would fall; if it was made of air, it would rise toward heaven. Moreover, Aristotle also deduced that an object twice as heavy as another object would fall twice as fast.

For more than a thousand years, people tried to classify objects according to their elemental nature. But then, near the end of the 16th century, Italian scientist Galileo Galilei asked himself what would happen if two stones of different mass were dropped at the same time from the top of the Leaning Tower of Pisa. When he did the experiment, Galileo discovered something startling: the two stones fell with the same acceleration and hit the ground at pretty much the same time. Since Aristotelian physics couldn’t account for this observational anomaly, Galileo’s experiment brought about a paradigm shift, after which future scientists started solving different kinds of puzzles. Namely, instead of classifying objects according to their elemental nature, they began trying to determine why objects with different masses should fall with the same acceleration. Everything changed: their worldviews, their assignments, their objectives. 

The most interesting aspect of new paradigms is that they are incommensurate – that is, inadequate – with old ones. In other words, you cannot be an adherent to Ptolemy and Copernicus simultaneously, because the Earth can either be the central rock of the universe or the third rock from the sun. Consequently, after a paradigm shift, the new paradigm – once established – completely supplants the old one. However, Kuhn remarks, some kind of paradigm (within which normal science would happen) must be present at all times. “The decision to reject one paradigm is always simultaneously the decision to accept another,” he writes. “To reject one paradigm without simultaneously substituting another is to reject science itself.” 

Final notes

One of “The Hundred Most Influential Books Since the Second World War” according to the Times Literary Supplement, Thomas Kuhn’s “The Structure of Scientific Revolutions” is, arguably, the defining work in the philosophy and history of science. 

“If you're making a list of books to read before you die,” wrote The Observer in 2012, “Kuhn's masterwork is one.” We tend to agree.

12min tip

“Truth emerges more readily from error than from confusion,” wrote Francis Bacon in the 17th century. So, treat your errors the way James Joyce did his: as portals of discovery.