In this blog post, we will examine the representative theories of Karl Popper and Thomas Kuhn, focusing on the demarcation problem—the distinction between science and nonscience—and explore the possibility of a new criterion for distinguishing science from nonscience that critically integrates these two perspectives.
The question of what criteria distinguish science from non-science is known as the “demarcation problem” and has long been one of the central themes in the philosophy of science. Numerous philosophers have sought to answer this question, and lively debate continues to this day. In fact, there is generally consensus among scholars regarding which fields are considered science and which are not. For example, physics, chemistry, and biology are classified as science, while astrology, religious doctrinal systems, and various forms of pseudoscience tend to be classified as non-science. However, there is still no single, widely accepted criterion regarding the fundamental basis for this distinction.
In this article, we will examine the arguments of the philosophers of science Karl Popper (1902–1994) and Thomas Kuhn (1922–1996), who have had a significant influence on the demarcation problem, and explore the possibility of establishing new demarcation criteria by critically revising and integrating their theories. In particular, we will examine how Popper’s concept of “falsifiability” needs to be combined with Kuhn’s concept of “paradigm,” while noting that, conversely, certain elements of Kuhn’s theory are not strictly necessary conditions for defining the essence of science. Based on this, we propose a new approach that defines science as “the activity of constructively investigating falsifiable theories within a paradigm.” Furthermore, we will examine whether this criterion can more effectively distinguish between actual science and non-science.
Popper proposed a simple yet powerful criterion regarding the demarcation problem. His central argument is that only falsifiable theories are science. At the time, science was often understood as a process of establishing true propositions through inductive reasoning that generalizes from observed cases. For example, if one observed a thousand sheep and found them all to be white, one might reach the conclusion that “all sheep are white.” However, such a conclusion cannot be logically established, as it is impossible to completely rule out the possibility that a black sheep exists in a place that has not yet been observed. Conversely, if even a single black sheep is discovered, the proposition “all sheep are white” immediately becomes false.
To address this problem with induction, Popper argued that the core of science lies not in inductive proof but in falsification through counterexamples. Science, he maintained, is a process of proposing falsifiable hypotheses and theories and then verifying, through repeated testing and examination, whether they are falsified. The fact that a theory survives does not mean it is absolutely true, but it can at least be considered to have passed the tests conducted so far. Therefore, Popper viewed the role of scientists as testing their theories as rigorously as possible.
However, recognizing falsifiability as a necessary condition for science is a separate issue from regarding it as a sufficient condition. In fact, while Popper cited astrology and certain psychological theories as examples of pseudoscience, contemporary philosophers of science point out that these fields are not considered non-scientific simply because they are unfalsifiable. Rather, the problem lies in the fact that, even in the face of repeated failed predictions, systematic revision and development do not take place. If a field were recognized as science solely on the basis of offering falsifiable predictions, Popper’s criteria would include an excessively broad range of disciplines under the umbrella of science.
That does not mean Popper’s theory must be discarded. Popper’s limitation lies in the fact that, by focusing too heavily on the logical structure of scientific theories, he failed to sufficiently consider how actual research communities develop and revise theories. For example, suppose astrology presents a prediction such as, “When the Sun and Jupiter form a specific angle, people born on a certain date are more likely to fall ill.” In principle, this proposition is empirically verifiable and falsifiable. Someone without scientific training might consider such a claim worthy of verification. Furthermore, if an “ideal astrology” existed that applied modern scientific research methodologies to continuously accumulate data and revise hypotheses, such an activity might appear quite similar to science in many respects.
However, astrology in reality has failed to systematically revise its repeatedly failed predictions or develop a cumulative body of knowledge. It is precisely for this reason that it is not recognized as science. This demonstrates that it is difficult to define science based solely on the logical structure of propositions. To understand science, one must also consider how researchers respond after a theory is falsified and how the field as a whole evolves.
Based on this line of thinking, Kuhn proposed a new perspective. In ‘The Structure of Scientific Revolutions’, he distinguished between “normal science”—conducted under a dominant paradigm—and “scientific revolutions,” during which paradigm shifts occur. According to Kuhn, the core of science lies not in individual theories but in paradigms, and researchers solve problems and accumulate knowledge within a shared paradigm.
In particular, Kuhn placed great importance on the activities of normal science. Normal science is the process of solving specific problems and refining theories based on the existing paradigm. Researchers generally do not question the basic premises of the paradigm, and even when discrepancies arise between observational results and theory, they first seek solutions within the paradigm itself. This contrasts with Popper’s view, which held that all theories must be continuously tested and sought to be falsified.
Kuhn’s explanation is significant in that it demonstrates how science actually progresses. He emphasized that the core of scientific progress lies not in revolutionary moments but in the everyday activities of normal science. In fact, most scientific achievements are accumulated during periods when the paradigm remains stable. Conversely, it is difficult to call a field “science” if it is characterized by a proliferation of theories that are unconnected to one another and lack cumulative progress. Science requires a consistent research tradition, a shared sense of inquiry, and productive research activities based on these foundations.
However, it is worth reexamining whether the element in Kuhn’s theory—that “researchers do not test the basic assumptions of the paradigm”—is absolutely necessary for defining the essence of science. Kuhn explained that, because real-world scientists conduct research within constraints of limited resources and time, most of them maintain the existing paradigm. However, this is merely a statement describing actual scientific activity; it is difficult to view it as a condition that defines the essence of science.
For instance, let us imagine an ideal environment. If sufficient personnel, resources, and time were provided, some researchers would strive to solve problems within the existing paradigm, while others could simultaneously conduct research to test the paradigm itself. For example, if retrograde planetary motion had been discovered during the era when the geocentric model was dominant, one group would have sought to resolve the problem by supplementing the existing system, while another group would have sought to reexamine the geocentric model itself.
Even in such a situation, both groups could be considered to be engaging in scientific activity. In fact, if diverse research strategies were pursued in parallel, a paradigm shift might occur more efficiently. Therefore, while the condition of not testing the basic assumptions of a paradigm is useful for explaining real-world science, it is difficult to view it as an essential condition for defining science.
Reflecting this point, Kuhn’s criterion can be simplified to “constructive research activity carried out within a paradigm.”
By combining Popper’s falsifiability with Kuhn’s revised criterion, we can arrive at the definition that science is “the activity of constructively investigating falsifiable theories within a paradigm.” In this definition, Popper’s falsifiability functions as a logical criterion for evaluating individual propositions and theories, while Kuhn’s concept of a paradigm functions as the social and research framework that allows such theories to develop within the actual academic system.
For this criterion to be valid, the two elements must first be compatible. The central debate between Popper and Kuhn centered on the extent to which researchers should rigorously test theories. However, if the intensity of testing is removed as a necessary condition from the very definition of science, the two positions can be fully reconciled. In practice, science has sometimes undergone the rigorous verification emphasized by Popper, and at other times has evolved in the form of “normal science” described by Kuhn. Therefore, these two elements are not mutually exclusive when defining the essence of science.
So, is falsifiability still a necessary condition? The answer is yes. A theoretical system that is unfalsifiable can still generate a certain amount of research activity. However, if its constituent propositions do not, in principle, allow for empirical verification, it is difficult to call such a system science. Science is connected to the empirical world and must acknowledge the potential for error in its claims. Falsifiability guarantees precisely this characteristic.
Falsifiability becomes even more important, particularly during paradigm shifts. Empirical verification and critical examination are essential for determining whether an existing theory actually has flaws and whether a new theory offers better explanatory power. Therefore, falsifiability plays a pivotal role in the maintenance and development of the scientific system.
Finally, let’s examine whether this criterion effectively distinguishes between science and pseudoscience. Science is based on falsifiable theories and, at the same time, conducts cumulative and constructive research within a paradigm. In contrast, pseudoscience or non-science often fails to meet one or more of these two conditions.
Taking astrology as a prime example, Popper criticized it for lacking falsifiability, while Kuhn criticized it for failing to establish a constructive research tradition characteristic of normal science. Although their interpretations differed, the criteria proposed in this paper require both conditions, allowing us to more clearly classify astrology as non-science. In this regard, the new criteria demonstrate the potential to partially address the limitations of existing criteria.
In conclusion, the demarcation problem is not merely a matter of academic classification but a crucial issue directly linked to what we should accept as reliable knowledge. Popper emphasized the logical aspects of science through falsifiability, while Kuhn explained the actual research process of science through paradigms and normal science. Although these two perspectives appear to conflict, appropriately combining their core elements allows for a richer understanding of the essence of science. The question of “What is science?” is becoming increasingly important in an era like today’s, which is overflowing with vast amounts of information and unverified claims. Furthermore, the answer to the demarcation problem can likely be found in the process of integrating the strengths of different philosophical perspectives.
