Demarcation: The Fine Line Between Science and Pseudoscience
Science is interesting but messy. It has flaws and limitations, but it remains our most effective tool to understand the natural world. Contrary to what many may think, science does not “prove” [1] anything and it does not lead to an “objective truth”. Science is the process by which scientists collect data through observational and experimental evidence to explain a particular phenomenon of the universe.
In light of new and growing evidence, explanations are updated as new data is gathered and as repeated testing increases our confidence in its certainty.
Scientific “facts” stem from testable and falsifiable hypotheses. They are explanations supported by robust evidence, with a high degree of certainty. However, even if a hypothesis is supported by a wealth of confirmatory evidence, it only takes one contrary observation to falsify it. On the other hand, an unfalsifiable hypothesis cannot be “proven” wrong or falsified by any scientific instrument. Falsifiability is a key aspect of science required for a claim to be considered scientific, at least in principle if not in practice [2]. The principle of falsifiability is considered a key aspect of the demarcation line between what is considered science and what is not.
Richard Feynman, on the Nature of Science
The Demarcation Line
A number of scientists and members of the public view pseudoscience, such as astrology and alternative therapies, as an area that clearly falls on the “not science” side of the demarcation line that divides science from pseudoscience. A bright line pointed out years ago by Karl Popper.
In his book, The Logic of Scientific Discovery, Karl Popper (1959) explains the principle of falsifiability in science as the “statements or systems of statements, in order to be ranked as scientific, must be capable of conflicting with possible, or conceivable observations” (p.39) [3]. Popper states that a valid hypothesis can be potentially falsified by empirical evidence and that scientific experiments should be designed in an attempt to falsify it, rather than verify or confirm it. A lack of falsifying evidence means that the hypothesis can be provisionally accepted as true unless new evidence emerges that suggests the opposite. In other words, further evidence that might falsify the hypothesis is always a possibility. Several hypotheses have accumulated enough strong evidence that they are unlikely to be falsified but the possibility, even if it is slight, is never ruled out.
Popper rejected verifiability as a criterion of science. One reason is that verifiability legitimizes existential statements as scientific, even though such statements cannot be falsified [4]. For example, someone might claim that unicorns exist. For this statement to be falsified, the alternative would be that unicorns do not exist. Science, by its very nature, does not provide evidence that proves the non-existence of anything.
Russell’s Teapot is the philosophical argument that illustrates that the burden of proof lies on the one making the claim, in this case, proving that unicorns exist. Believing that unicorns exist, even though there is no evidence, is supported by the mere fact that we have failed to see a unicorn, yet. On the other hand, the criterion of falsifiability renders such statements unscientific because they cannot be falsified. How can science ever prove that unicorns do not exist?
Falsifiability in Pseudoscience
In the philosophy of science, the demarcation problem differentiates science from parascientific domains, non-scientific domains that are not pseudoscientific such as history, philosophy, art, or religion. More importantly, it examines the boundaries between science and pseudoscience. Unfalsifiable claims, common in pseudoscience, fall in the realm of irrational discourse and lack supporting evidence and reason. While pseudoscience looks for evidence to support a pre-defined claim, science is constantly challenging claims and looks for evidence that might falsify the claim. In other words, pseudoscience pursues confirmation and science seeks falsification. Scientists can perform tests that might falsify a scientific claim, but no conceivable test could show a pseudoscientific claim to be false.
Falsification as a scientific practice is particularly significant because we like certainty. We are inclined to look for evidence that supports rather than contests our existing opinions, a phenomenon known as confirmation bias. For example, an astrologist would explain that Geminis are indecisive and extroverted and would continue to support their beliefs by giving examples of celebrities or politicians with such characteristics. Proponents of astrology, through confirmation bias, would then tend to see validations of this theory everywhere.
Falsifiability, and thus testability, is a crucial starting point for designing experiments to explain the phenomenon under study. If significant results are obtained, then the falsifiable theory would become accepted. Popper explains that scientists make claims and gather evidence to falsify the claim. If they fail, then the claim remains in effect, under the condition that scientists are always on the lookout for falsifying evidence. According to Popper, experiments might provide evidence that establishes, with certainty, that a claim is false. However, we can never establish, with certainty, that a claim is true. Hypotheses and theories remain in the tentative stage. This is not to say that science never resulted in false claims. On the contrary, science attempts to discard false claims, an approach that does not occur with pseudoscience.
Misinformation and Falsifiability
Science is our most reliable source of knowledge in a wide range of areas. Its high status often leads to exaggerated claims, which leads to misinformation surrounding pseudoscience that elevates its unfounded credibility. This gives rise to ineffective and sometimes dangerous interventions making the demarcation issue pressing in many areas. As a result, most countries have at least one professional organization dedicated to identifying, debunking, and combating pseudoscience. Examples include parents who refuse to vaccinate their children, people who pay large amounts of money to pseudoscience-peddling charlatans, climate denialism that hampers political action, and the myths associated with genetically modified food, to list a few.
Falsifiable hypotheses provide one of the following three outcomes:
Results reject the hypothesis
Results support the hypothesis/fail to disprove the hypothesis
Results are inconclusive
Unfalsifiable hypotheses provide one of the following two outcomes:
Results support the hypothesis
Results are inconclusive
Therefore, an unfalsifiable hypothesis cannot be disproven, and that is not because it is a sound, scientific hypothesis but because for any test you run, the hypothesis cannot be rejected or falsified. Results only provide inconclusive evidence.
Example: Vaccines contain microchips.
Ideally, the statement is falsifiable as scientists could run tests and not find any microchips. Someone could then argue that the microchips are so small that no scientific test can detect them. In this case, the results did not reject the hypothesis but are instead deemed inconclusive. For those who believe that vaccines contain microchips, the hypothesis is unfalsifiable and negative results only fit the narrative. For those who engage in conspiratorial thinking, the lack of evidence is evidence that their claim is the “truth”.
Conjectures and Tentative Theories
Popper’s notion of falsification is not the only or the primary sign of science, but it can be used to evaluate new ideas and decide how much weight to give them. Provocative conjectures regularly circulate in the media and experts are entrusted with the task of carefully assessing their credibility. The COVID-19 pandemic offered several examples of conjectures that received irrational credibility by the public even though they were not supported by rigorous evaluation, often leading to patient harm.
Such conjectures about medicine were the result of anecdotal evidence, uncontrolled clinical trials, or animal models, all of which are not trustworthy and have a high risk of bias. One example is hydroxychloroquine, which was hyped as a breakthrough treatment for COVID-19 on the basis of in vitro studies demonstrating anti-viral activity. Prescriptions of hydroxychloroquine surged, despite the hypothesis being falsified by multiple clinical trials showing no benefit, and possibly increased harm [5].
An experiment might fail to support a hypothesis. This does not mean the hypothesis is wrong because it was ‘falsified’ by the experiment. In fact, any element of the set-up could be the cause. To test a hypothesis, scientists make several supporting hypotheses, assuming that the experimental setup lacks any flaws, the math is correct, and all variables are accounted for in the experimental design. In this case, when the result falsifies the original hypothesis, one can investigate whether the error lies in the original hypothesis or the supporting hypotheses.
Unlike conjectures, tentative theories emerge from volumes of data gathered from years of scientific studies . Theories, however, are complex and the binary concept of falsifiability does not strictly apply. Theories might be supported under certain conditions, rejected under other conditions, or partially supported depending on the variables considered. Such nuances should be considered to resist the tendency to oversimplify the concept of falsifiability. Theories supported by a wealth of evidence, over a long period of time, and tested repeatedly using high-quality studies, have withstood Popper’s falsifiability and serve as a basis to evaluate new evidence. For theories that have been falsified or corroborated, it is important to identify what tests were performed and if the results falsified certain aspects of the theory or the whole theory.
Conclusion
Many of our decisions depend on our understanding of science; therefore, drawing the line between what is science and what is not is important to avoid being deceived by pseudoscientific claims that offer the same level of authority as science. It might be appealing to seek a single principle that we can use to understand and apply the essence of science. Unfortunately, science is more complicated than that. It is not a linear, structured process of conjecture and refutation. It is the ability to question observations, assumptions, and hypotheses and in doing so, unfold new possibilities that were previously hidden.
References
[1] Geraint , Lewis. “Where's the Proof in Science? There Is None.” The Conversation, 23 Sept. 2014.
[3] Popper, Karl. The Logic of Scientific Discovery. Basic Books, 1959.