Article Review: The Idea That a Scientific Theory can be ‘Falsified’ is a Myth
by Mano Singham
The Idea That a Scientific Theory can be ‘Falsified’ is a Myth is an article by Mano Singham that appeared in Scientific American on 7 September 2020. It discusses a key feature of Karl Popper’s philosophy, his theory of falsification. Popper grew up in Vienna in the early 20th century and was in contact with the Vienna Circle but was opposed to their attempts to create a theory of confirmation. For Popper, theories could never be confirmed, as he believed that our scientific knowledge is always provisional. Popper proposed the theory of falsification as a means of demarcating science from non-science. It posits that for a theory to be considered scientific, it must propose testable predictions that are able to be proven false. Popper’s theory of falsification has since become so popular that it is now widely considered a foundation of the scientific method. But this is exactly what Singham’s article is reacting against.
Singham argues that the philosophical, sociological, and historical study of science has shown “that falsification cannot work even in principle…because an experimental result is not a simple fact obtained directly from nature.” Here, Singham seems to be referring to several issues discussed in the philosophy of science: natural kinds, theory-ladenness of observation, and confirmation holism.
Natural kinds is the idea that science is assumed to be identifying the objective structure of nature and separating nature into individual units and correctly identifying these individual units as different kinds. However, it may be that these different kinds are not separate individual units but simply seem to be due to our own conceptual understanding of nature. That is, it may be that nature has no objective structure, and that the different kinds we categorize as separate individual units are differences for us, but not necessarily reflective of any objective difference in nature. The problem of natural kinds demonstrates an issue for any theory of confirmation or falsification, since it is difficult to know with certainty whether these different natural kinds even exist, let alone are being correctly identified.
Theory-ladenness of observation is the idea that it is our own theoretical understanding that influences how we interpret any observation. That is, in every observation we make, we need to utilize the kinds of terminology and measurements that have been adopted by a scientific community to interpret them. Therefore, it is impossible for anyone to understand any observation without first understanding the theories that underpin those observations. This inability to disentangle observations from the theories that underpin them also demonstrates an issue for falsification.
Confirmation holism is an idea that was originally proposed by Willard Van Orman Quine in his paper Two Dogmas of Empiricism as an argument against a theory of confirmation, but it serves as an argument against Popper’s theory of falsification too. Confirmation holism is the idea that it is impossible to test any claim in isolation of an entire holistic understanding of the world. That is, whenever one seeks to test any one claim there are a vast number of assumptions that are being made which underpin that claim. Therefore, any one test of an idea is actually a test of a massive conjunction of ideas, where any failure of the test does not necessarily indicate a failure in the specific idea being tested, since it could be a failure anywhere in that massive conjunction of ideas. The problem of natural kinds, the theory-ladenness of observation, and confirmation holism is likely what Singham means when he says “we cannot say that any particular theory is falsified.”
Singham then moves on to say that “falsification—or any other philosophy of science—is not necessary for the actual practice of science” because “actual scientific history reveals that scientists break all the rules all the time.” He mentions an argument by Thomas Kuhn as an example, whose influential work The Structure of Scientific Revolutions demonstrated that scientists don’t just give up their ideas as easily as Popper’s theory of falsification would suggest. Kuhn, along with other critics of Popper, such as Imre Lakatos and Paul Feyerabend, all rejected the idea that there is one single methodological rule that can be applied to all of science and account for its progress. Philosophers of science such as this painted a picture of science that was largely sociological.
“…the knowledge gained from these fields allows science advocates to be better equipped against forces that aim to undermine science…”
Singham laments that this sociological picture of science has caused some scientists to criticize these external studies of science, since they can be seen as undermining the objectivity of, and public confidence in, science. Singham argues against this on two fronts. Firstly, that these studies of science give a richer understanding of the field that would inspire “more confidence in science, not less.” But, without any empirical evidence to support the idea that these fields do inspire confidence in science (and that it is not just those that already have confidence in science who decide to study these topics), this claim appears to be purely speculative. Secondly, he claims that the knowledge gained from these fields allows science advocates to be better equipped against forces that aim to undermine science. This is likely to be true, since those who are already confident in science are going to be better equipped to make persuasive arguments to support their case after studying fields such as these.
Singham lists climate denial, tobacco, anti-vaccination, and anti-evolution advocates as examples of how the prevalent idea of falsification in science has been used as a tool to undermine science. He argues that these anti-science advocates fund and/or conduct their own scientific studies with the intention of producing results that conflict with the scientific consensus in order to create doubt and allow them to “argue that they have falsified the consensus.” Here, Singham is making several arguments at once.
Firstly, the suggestion that it is the scientific consensus that matters is aligned with Helen Longino’s solution to the idea that the sociological analysis of science presents scientific knowledge as a social construct, undermining its objectivity. But Longino argues that it is the very social structure of science that allows it to create objectivity. She states that there is mutual agreement between scientists on the best methodology to achieve objectivity, and that the peer-review process acts as a kind of conventionalized intersubjectivity where personal bias is systematically removed, leading to even greater objectivity. Therefore, according to this view, once there is high agreement amongst a scientific community, it is a good indicator that the science is sound. This picture of science is embraced when Singham says that it is the “comprehensive bodies of evidence to arrive at consensus judgments about whether a theory should be retained or rejected in favor of a new one.”
Secondly, while I do believe Singham’s assertion that anti-science forces seek to undermine science using the aforementioned techniques, his argument that it is the idea of falsification that allows them to do this is rather dubious. This is because these techniques would have the same impact of increasing doubt whether the idea of falsification was prevalent or not. It are cognitive biases (such as confirmation bias) and a lack of scientific literacy that allows this impact to happen, not the idea of falsification. So, while falsification is fundamentally flawed and is certainly not the foundational principle of science that Popper claimed it to be—or even applicable to all scientific disciplines—it still proves to be a useful tool for scientists to implement when appropriate.
