This blog post compares the core concepts of these two thinkers, examining how science changes and develops from various perspectives. Let’s explore together the dynamism of science revealed by these differing theories.
Throughout my 12 years of schooling leading up to university, I had always regarded science as an objective, absolute, and unconditionally trustworthy discipline, never deeply contemplating its essence. This is likely true for the majority of students who have followed a similar educational path. However, before studying the detailed theories and principles of science, we must first consider what the essence of science truly is. Just as we study history to explore future directions based on past facts, science requires deep reflection on the history of its philosophy to ensure its continuous and correct development. Philosophers’ views on scientific progress diverge significantly, with Popper’s falsificationism and Kuhn’s paradigm theory being representative examples. Both theories possess distinct strengths and weaknesses. This paper aims to expand Kuhn’s paradigm theory based on Popper’s falsificationism.
Chronologically, Popper’s falsificationism precedes Kuhn’s paradigm theory. To address Popper’s falsificationism, it is necessary to first examine the inductivism of the preceding era. Statements derived from observation are called observational statements, and inductivists argue that universal laws can be deduced from them. Specifically, they state that if a phenomenon consistent with a particular universal statement is sufficiently repeated under various conditions, and not a single contradictory observation is found, then generalizing that law is justified. For example, if it has been observed multiple times that metal always expands when heated, regardless of the type of metal or the method of heating, and there has never been a single instance of contraction, an inductivist would consider it possible to derive the universal statement: “Metal expands when heated.”
However, this inductive approach faces several problems. First, the criteria for what inductivists mean by “sufficiently many” samples and “various” conditions are unclear. It is impossible to know how many experiments are needed to justify a universal statement. Furthermore, while the population obtainable from experiments is generally infinite, the actual sample that can be investigated is finite. According to classical probability theory, no matter how many experiments are conducted to confirm a universal statement, the probability that the statement is true converges to zero. This means that conducting more experiments does not increase the likelihood that a statement is true. Furthermore, observations are influenced by the observer’s knowledge, experience, expectations, etc., and can therefore appear different depending on the observer. That is, theory always precedes observation, which contradicts the inductivist view that laws are proven through the accumulation of observations. Thus, inductivism has significant limitations. Unlike this inductivism, Popper rejected the position that science is based on induction and proposed falsificationism.
Falsificationists acknowledge that theory precedes observation and emphasize the importance of conjecture and refutation in science. According to falsificationism, science progresses through the process of proposing theories and refuting them, and only theories that survive this process endure. As confirmed earlier in inductivism, proving a theory to be true within finite time through experiments is impossible. Conversely, revealing a theory to be false is relatively easy, as it only requires finding a single contradictory observational statement. Falsificationists explain scientific progress based on this property: hypotheses are falsified, improved upon in the process, and science advances. This also means that hypotheses impossible to falsify are scientifically meaningless. For science to progress, newly proposed hypotheses must be more susceptible to falsification than existing ones. For example, the hypothesis “All planets in the solar system orbit the sun in elliptical orbits” is more advanced than the hypothesis “Mars orbits the sun in an elliptical orbit” because it is more susceptible to falsification.
The set of theories generally accepted in a given era constitutes the background knowledge of that era. The more a conjecture diverges from the background knowledge, the bolder it is; the more it aligns with the background knowledge, the more cautious it is. To make a meaningful contribution to scientific progress, bold conjectures must be confirmed, and cautious conjectures must be falsified. While the ‘confirmation’ invoked by inductivists remains confined to the probabilistic and logical relationship between observational statements and universal statements, confirmation in falsificationism carries the meaning that a theory successfully predicts new facts that contradict existing background knowledge.
However, falsificationism also has limitations. As mentioned earlier, observational statements themselves can be erroneous, and since observations are preceded by specific theories, completely certain falsification cannot exist. Furthermore, historically, there exist cases that falsificationism cannot explain. For instance, the transition from the geocentric to the heliocentric model took over 100 years, and the existence of a single falsifying case does not immediately lead to the immediate abandonment of an existing theory. Thus, both inductivism and falsificationism are overly simplistic to explain the actual process of scientific growth. To compensate for these limitations, Kuhn’s paradigm theory emerged.
According to Alan Chalmers, Kuhn’s paradigm refers to the set of assumptions, laws, and techniques shared by scientists in a particular era. Research conducted within this paradigm is called normal science. Within normal science, scientists clarify and extend the scope of the paradigm’s application. Suppose an anomalous case emerges that cannot be explained by the existing paradigm. Popper’s falsificationism holds that such an instance constitutes a single falsification of the paradigm, demanding its immediate abandonment and replacement with a new paradigm. However, Kuhn’s explanation states that a paradigm is only considered falsified when multiple anomalous cases accumulate, leading to a crisis. This effectively explains the transition from geocentrism to heliocentrism, which falsificationism could not account for. When a paradigm is falsified, scientists adopt a new paradigm, and a new period of normal science begins. Kuhn calls this process a scientific revolution.
According to Kuhn, normal science is governed by a single paradigm; two or more paradigms cannot coexist within the normal science of a single era. That is, unlike inductivists, Kuhn does not view science as developing through the accumulation of observations. Instead, he sees science as alternating between periods of normal science and scientific revolutions, during which paradigms are ‘replaced’. However, this perspective contradicts the actual cumulative growth of science when viewed macro-historically. While fundamental theory replacement, as Kuhn argues, occurs over centuries, during the same period, diverse paradigms have taken root more deeply and broadly across various scientific domains, clearly demonstrating cumulative development of scientific theory. Therefore, the author seeks to realistically expand Kuhn’s paradigm by utilizing Popper’s falsificationism.
As demonstrated in the examples below, I contend that the set of paradigms across scientific domains changes primarily through three modes: modification, generation, and integration. Scientific progress can be understood by how much the set of paradigms expands the ‘size of the explainable domain’. Here, ‘explanation’ signifies not merely the absence of contradiction but the possibility of direct confirmation.
First, paradigms can change. This resembles the phenomenon described by Kuhn. For a paradigm shift to occur, falsification sufficient to create a crisis within the existing normal science is required. This falsification arises when a new paradigm can explain phenomena that the old paradigm cannot. A new paradigm that fails to explain phenomena the existing paradigm can explain already contains refutation and thus cannot be accepted. That is, it lacks the explanatory power to replace the existing paradigm. Ad hoc explanations are excluded here.
Second, paradigms can be newly created. The primary theory constituting a paradigm is called the fundamental theory, and theories derived from it are called subtheories. Fundamental theories provide new tools and frameworks for thought. For example, advances in optics enabled the microscope, which led to the birth of new fundamental theories like cytology and created a new paradigm. Expansion into new domains broadens the scope of what science can explain.
Third, paradigms can be integrated. The unification of electric and magnetic fields is a prime example. An integrated paradigm enables explanations not only for the domains previously covered by existing paradigms but also provides additional explanations inherent to the integration itself.
Thus, as the set of paradigms expands the scope of what can be explained through their change, creation, and integration, science progresses. The notion that a paradigm with a broader explanatory scope represents a more advanced theory aligns with Popper’s falsificationism. A broader explanatory scope inherently implies a greater potential for falsification. The ‘explanation’ in Popper’s ‘theory that can explain everything’ differs in nature from the ‘explanation’ discussed here. Popper refers to a theory that is logically unfalsifiable, whereas the explanation in this text includes the possibility of direct confirmation.
Popper’s falsificationism resolved most problems of inductivism, but it failed to account for the complexity of actual scientific development because it demanded the abandonment of theories regardless of the degree of falsification. Conversely, Kuhn’s paradigm theory could explain the process by which scientific theories change over long periods, yet it did not fully reflect the cumulative development of actual scientific theories. Therefore, the author focuses on how science’s explanatory domain expands through paradigm shifts, generation, and integration. The author seeks to define scientific progress as the size of the domain explainable by a set of paradigms. This is fundamentally aligned with Popper’s falsificationism, as the explanatory domain’s expansion inherently increases falsifiability.
