In this blog post, we will look at how inductivism and empiricism tried to establish scientific knowledge and the limitations of these theories.
Scientific philosophy has developed by presenting a method for establishing scientific knowledge as knowledge through a logical procedure. Inductivism, which was advocated by Francis Bacon and others in the 17th century, played an important role in the establishment of modern scientific philosophy. Scientists of this period believed that the inductive approach, which involves collecting empirical data and deriving general rules from it, was the best way to discover scientific truth. In the early stages of scientific inquiry, inductivism emphasized the importance of experimentation and observation in particular, and was seen as an essential tool for scientists to expand their understanding of natural phenomena.
The subsequent disconfirmation theory sought to address the logical problems of inductivism. Falsificationism was proposed by Karl Raimund Popper, who provided a new methodology for verifying the truth of scientific theories. Falsificationism emphasizes that all scientific knowledge is temporary and can be modified whenever new evidence is presented. Therefore, falsificationism claims that the development of science is achieved through repeated falsification and modification of theories. However, has falsificationism completely solved the problems of inductionism?
Inductionism is a type of logical structure that obtains scientific knowledge through induction, where induction is the creation of a certain rule empirically based on various observations. If a certain phenomenon is repeated without exception, it is assumed that the remaining cases will be the same as the observed cases based on the examples. It was Francis Bacon who organized this into a logical step. In the 17th century, Francis Bacon proposed a premise for obtaining scientific knowledge: First, when collecting examples, one should not have previous beliefs. Francis Bacon, for example, directly criticized the scientific knowledge of Aristotle, which had been believed in Europe for about 2,000 years. Next, to obtain the desired facts, we observe a specific situation, and we use methods such as experiments. Finally, the observed results are generalized, that is, scientific theories are derived using induction. When science is explained in this way, it can be considered to be quite reliable because it is explained based on objective data.
However, there is a fatal flaw in the method of induction, which is the creation of scientific knowledge through objective data, and that is that the principle of induction itself cannot be logically justified. To give a famous example, let’s say that through a lot of observation, we have gained scientific knowledge that “all swans are white.” This knowledge was inductively obtained by observing individual swans and confirming that all observed swans are white. However, just because all swans observed so far are white does not guarantee that all swans that will be observed in the future will be white. To obtain this knowledge, not all swans in the world have been observed, so the object of the black swan may not have been observed yet. In other words, it may be incorrect knowledge due to a lack of observation. As can be seen in this example, the inductive method is a method of explaining the whole by combining individual objects, so the method of deriving results is inevitably incomplete.
The limitations of inductive reasoning are not limited to simply logical problems. As scientific discoveries progress, it frequently happens that previous generalized laws are invalidated by new evidence. For example, in classical mechanics, Newton’s laws were long considered scientific truths, but the emergence of Einstein’s theory of relativity revealed that these laws could only be applied under special conditions. These historical examples clearly show the imperfection of inductivism. Nevertheless, inductivism played an important role in the early stages of scientific inquiry by systematically analyzing and generalizing empirical data.
Therefore, Karl Raimund Popper advocated falsificationism to solve this problem of induction. Falsificationism is a scientific view that a theory is the best theory if it survives repeated attempts to refute it. According to Karl Raimund Popper, scientific theories are always characterized by the fact that they can be refuted by experienced facts, which is called the requirement of refutability. Therefore, no theory can be considered true just because it has not been disproved. However, such theories can be considered superior to previous theories. In other words, the theory of falsificationists holds that no scientific theory is completely true. They believe that theories in a particular field continue to progress as they become increasingly superior, with theories that withstand multiple refutations gradually emerging.
As can be seen above, counterfactualism has solved to some extent the problem of inductionism, which claims that a theory is not true if it is not true in every individual case. The fact that the theory of counterproofs emphasizes that science is a process of constant revision and development rather than a set of definitive truths has deepened our understanding of the nature of scientific inquiry. However, the theory of counterproofs has not solved the fundamental problems of inductionism, which are as follows.
First, the results of observation can be wrong. In the theory of refutationalism, when refuting a theory, the method of refuting the existing theory by observing counterexamples was adopted. However, when conducting observations or experiments to find counterexamples, the existing theory is used, which, from the perspective of Karl Popper’s theory of refutationalism, are the best theories currently available, but they are not true. In other words, there is no guarantee that a counterexample is true, so counterexamples cannot properly refute existing theories. In order to refute a particular theory, the observation that refutes it must be based on some theory, but according to refutationalism, all theories are not true, so there is no basis for refuting a particular theory. In inductivism, logical leaps occur due to inaccuracies in observation, and the theory of counterproofism was created to replace this. Counterproofism does not solve this problem at all.
Second, the part that is the subject of the counterproof is unclear. French philosopher of science Gilbert Durand says that scientific propositions are not independent but are combined with auxiliary assumptions and can be verified. In other words, if T is a theory, O is an observational statement, and A is an auxiliary assumption, then it is not only T that implies O, but T&A. Van Quine says that mathematical statements and logic can also be included in the place of A. Therefore, when testing a specific theory, it is impossible to determine whether theory T is wrong or A is wrong. To refute a specific theory, one must find out that T is wrong, but in fact, T&A may have been derived incorrectly due to A’s mistake. Therefore, the method of refutation used by refutationalism cannot be used to be certain that a specific theory is wrong.
Third, the theory of counter-evidence has limitations in explaining the progress of scientific theories. The theory of counter-evidence claims that a theory can be temporarily accepted if it is not disproved, but this does not sufficiently explain scientific revolutions or paradigm shifts. For example, Thomas Kuhn’s theory of paradigms shows that scientific progress is made by leaps and bounds rather than gradually, and that in this process, past theories are completely discarded and new theories are accepted. These examples suggest that the theory of falsification does not fully capture the actual way science progresses.
For these reasons, the theory of counterproofs was also evaluated as not being a suitable theory for verifying scientific knowledge. Not only did the theory of counterproofs fail to solve the problems of induction, but it also had its own problems. Considering this, the theory of counterproofs was not a good substitute for induction at all.
