This blog post critically examines whether Karl Popper’s falsificationism can truly overcome the limitations of inductivism.
Since the 17th-century Scientific Revolution, scientists have regarded empirical facts—such as experimental results and observations—as the foundation of knowledge. This led to the emergence of inductivism, a method of reasoning that derives general conclusions capable of explaining specific facts or events. However, Karl Raimund Popper, in his book ‘The Logic of Scientific Discovery’, pointed out the limitations of inductivism, arguing that no observation can precede a hypothesis or theory. He proposed falsificationism as an alternative to replace it. Popper argued that hypotheses and theories are proposed through a process of conjecture and refutation, and are accepted as tentative facts through attempts at falsification. This paper aims to demonstrate that Karl Popper’s falsificationism cannot serve as an alternative to inductivism because it shares similar problems with inductivism.
Karl Popper identifies the theory-dependence of observation as one problem with inductivism. According to Popper, some kind of theory always precedes any observation. He argues that observational statements, like theories, are fallible and thus cannot provide a solid foundation for supporting scientific theories and laws. Therefore, he contends that induction—deriving general conclusions from observational results—must inevitably be wrong.
Falsificationism was proposed as an alternative. Falsificationism is the perspective that science progresses as hypotheses or theories are continuously subjected to attempts at falsification through observation or experiment, and falsified hypotheses or theories are replaced by superior ones. Falsificationists argue that scientific hypotheses or theories must be falsifiable, and that these hypotheses or theories become increasingly superior as they overcome attempts at falsification. The position of falsificationists is that the more attempts at falsification a hypothesis or theory overcomes, the more it is accepted as a provisional fact, but it can never be definitively established as true.
However, despite being presented as an alternative to inductivism, falsificationism shares the same problems with inductivism that Karl Popper pointed out. First, complete falsification cannot be achieved in falsificationism. Since falsificationism holds that no theory can be definitively established as true, the theory itself is incomplete as a basis for falsification. Therefore, other observations must serve as the basis. However, due to the theory-dependent nature of observation explained earlier, observation cannot provide a solid foundation supporting theories and laws. Ultimately, the problem Karl Popper identified in inductivism manifests identically in falsificationism, leading to the conclusion that hypotheses or theories susceptible to falsification cannot be completely falsified.
Second, the manner in which hypotheses or theories develop, as presented by Karl Popper, is not significantly different from how they develop in inductivism. Popper argues that the process of theorizing in science is achieved through conjecture, and these conjectures are challenged by refutation through observation and experimentation. If these conjectures are proven false by observation, the conclusion is discarded. Karl Popper’s theory of conjecture and refutation posits that through trial and error, falsehoods are eliminated and truth is reached. Here, conjecture is not merely a collection of experimental results, but a bold conjecture open to refutation. Let’s take the example of crows. Seeing ten black crows while walking and proposing the hypothesis “Crows are black” is induction. Having the same experience and proposing a rule through conjecture—“Crows will be black”—is the method of conjecture. I believe there is actually little difference between the two methods. The characteristic of conjecture is that it can be falsified and, if proven false, discarded, which is similar to induction. Discovering a non-black crow refutes the former hypothesis; that hypothesis is judged false and discarded. Furthermore, conjecture is fundamentally based on observation, just like induction. Conjecture and induction are similar; even if they differ, since conjecture is based on observation, it is difficult to view the process of conjecture and refutation as superior to induction due to the theory-dependent nature of observation.
For these reasons, I cannot agree with the claim that Karl Popper proposed falsificationism to overcome the limitations of inductivism. As mentioned earlier, induction, refutation, and falsification all rely on observation and thus cannot escape the limitations of observation. Rather, I believe it is more valid to view falsificationism as a form of inductivism. Fundamentally, in falsificationism, conjecture, refutation, and falsification all derive conclusions based on observation. Given that induction is defined as a method of reasoning that draws general conclusions to explain specific facts or events, if falsificationism is regarded as critical induction and incorporated into induction, induction could be further developed. Assuming complete falsification is possible, critical inductivism holds greater validity than traditional inductivism because it allows the rejection of false theories through falsification and enables the evaluation of existing hypotheses or theories based on their falsifiability. Of course, even if falsificationism is incorporated into inductivism, the problem of the theory-dependence of observation remains, so the limitations of inductivism persist. Therefore, the process of proposing and evaluating theories through induction and falsification can still be considered incomplete and non-definite. To advance science, I believe a new claim capable of resolving these fundamental problems of observation and the theory-dependent nature must be presented.
