Popper and Kuhn on Demarcation

By Sriram Thothathri

Susan Miller, a prominent astrologist, stated on CBS New York that 2020 would “be a prosperous year” and that certain periods of time would be full of wealth and travel (Phelan). However, as shown by the impact of COVID-19 on the world economy and travel, this prediction turns out to be less than adequate. This prediction also brings into light the question of whether astrology is as credible and scientific as it claims to be. How can we determine whether astrology is a science or a pseudoscience? Philosophers of science Sir Karl Popper and Thomas Kuhn weigh in on this question of demarcation of science from pseudoscience with seemingly polar viewpoints. Popper argues that falsification and test-criteria should be at the core of demarcation, whereas Kuhn argues that the puzzle-solving nature of science should be at the core of demarcation. In this essay, I will argue that, while Popper builds a compelling distinction between verification and falsification to support his demarcation criteria, Kuhn incorporates the history of science and builds a more representative demarcation criteria through his discussion of the puzzle-solving nature of science.

Popper posits that falsification and testability should be at the core of demarcation as opposed to verifiability. Verifiability is measured by a theory’s propensity to be supported by observations and further experiments. Popper states that he, “could not think of any human behavior which could not be interpreted in terms of [pseudoscientific] theories” (Popper 1963/2012, 6). According to Popper, the unwavering verifiability of the theory should raise alarms as to the scientific nature of the theory at hand because pseudoscientific theories can be just as, if not more, verifiable than scientific theories. 

When I open my favorite astrology app on my phone, I often find that the remarks and affirmations it provides me with are deeply tailored to my life experiences. These affirmations often force me to recall verifying life experiences. For example, if my chart says that my health and wellness charts are maligned this month, I may verify this chart by referring to my gluttonous consumption of ice cream. On the other hand, if my chart says that I will be healthy, I may recall my frequent exercise. In both cases, the accessible evidence does not change, but the interpretation of existing evidence changes (usually to fit the affirmation). The astrological theories the app gave to me were usually flexible in that they stretch to fit many of my observations about myself, but Popper would likely note that these theories do not state anything scientifically meaningful as they are often overly generalized. In other words, a theory that seems verifiable by all evidence is not inherently a scientific theory.

Hence, Popper proposes using falsifiability, rather than verifiability, as a benchmark for the scientific nature of a theory. Falsifiability, in contrast to verifiability, describes a theory’s ability to be disconfirmed by observations.  For instance, Einstein’s gravitational theory stated that light is attracted to heavy objects such as the sun. Not only does this theory imply that light should curve when moving around massive objects that warp the spacetime around them, but it also predicts that distant stars should appear shifted between photographs of stars taken at night or during an eclipse. Einstein made a risky prediction because he set rigid standards by which he could observe certain phenomena that would render his theory false. On the other hand, if his theory is largely supported by the results the scientific community generates, the risk he took would have been fruitful and his research expands the reach of scientific knowledge. 

In Popper’s words, a scientific theory is, “incompatible with certain possible results of observation” (Popper 1963/2012, 6). Thus, a scientific theory can make risky predictions and has the potential to be falsified, while a pseudoscientific theory often does not set falsifying conditions and simply makes the rules as it goes along. That is not to say that scientific theories should not be modified to account for new phenomena. Popper argues against ad-hoc adjustments to theories which, like pseudoscientific theories, tend to overfit the phenomena to the data. To re-interpret a scientific theory without diminishing the theory’s scientific nature, the theory must fit Popper’s definition of a risky prediction. So, the theory must provide conditions such that the theory is disprovable with certain observations. In other words, adjustments to a scientific theory must allow for risky predictions to be made. These risky predictions take a risk of being false, but, if evidence generally supports the new theory, the theory would have said something new about a phenomenon and grown our scientific understanding of the phenomenon. In other words, adjustments to a scientific theory must allow for risky predictions for the theory to be scientifically valid and fruitful. 

While Popper focuses on testing and falsification to demarcate science from pseudoscience, Kuhn suggests that the puzzle-solving nature of science is a better measure of demarcation. To articulate Kuhn’s argument, I must define two terms: extraordinary science and normal science. Extraordinary science is defined by intermittent growth through, “overthrow of an accepted theory and its replacement by a better one” (Kuhn 1970/2012, 12). Normal science is characterized by scientists using existing theory to define a puzzle and, “guarantee that, given sufficient brilliance, it can be solved” (Kuhn 1970/2012, 12). In short, normal science is not focused on attempting to break every paradigm; instead, it attempts to solve puzzles in existing paradigms. Through this framework, Kuhn postulates that a focus on falsification is more characteristic of extraordinary science.

To illustrate the distinction, consider the puzzling discovery of a new planet. By the mid-eighteenth century, astronomers had noticed some discrepancies between the data regarding the orbit of Uranus and Newton’s laws of gravitation. French astronomer Alexis Bouvard reviewed this data extensively, but he was unable to determine whether these discrepancies rose from the inability of Newton’s laws to explain Uranus’s orbit or whether they were a result of some, “foreign and unperceived cause” (O’Connor). By Kuhn’s logic, the astronomer performing normal research should not immediately assume that they discovered a paradigm-breaking anomaly. Rather, the astronomer should try to solve the puzzle at hand by analyzing the relevant variables, instrumentation, calculations, et cetera. At the end of this puzzle-solving enterprise, the geologist must have either “resolved the puzzle he had been set” or “[abandoned] the puzzle entirely” in favor of using new hypotheses to re-evaluate the puzzling phenomena (Kuhn 1970/2012, 11). Fortunately, successive astronomers re-evaluated this phenomena with a new hypothesis that the “foreign and unperceived cause” was a new planet, and they resolved this puzzle by uncovering data consistent with both a new planet and Newton’s laws of gravitation (O’Connor). You might know this planet by the name of Neptune. In this case, there is no immediate assumption that existing theory is wrong and there is no room for critical discourse regarding existing theory. When Popper emphasizes falsification and risky predictions, Popper implies that there is room for critical discourse in science as risky predictions often go against existing theory. In this case, critical discourse means to test or criticize existing theory.

Kuhn flips Popper’s view on its head when he states that it is, “the abandonment of critical discourse that marks the transition to science” (Kuhn 1970/2012, 14). Imagine you have a jigsaw puzzle. To Kuhn, allowing critical discourse in the more representative normal science is like making new pieces or breaking existing pieces to solve a solvable jigsaw puzzle. Thus, critical discourse should only be allowed when the puzzle itself has irreconcilable problems. These irreconcilable problems can be likened to paradigm-breaking anomalies during periods of extraordinary science. All in all, Kuhn proposes an alternate method of demarcation by considering the abandonment of critical discourse and the puzzle-solving nature of a given field.

This does not mean existing theory should never be criticized. Rather, existing theory must be explored and built out sufficiently until the existing theory leads to contradictions where it appears that the puzzle is seemingly broken. Thus, critical discourse in science only occurs in periods of crisis or when, “the bases of the field are again in jeopardy” (Kuhn 1970/2012, 14). When applying this framework to astrology, Kuhn argues that, while astrology has been historically capable of making risky predictions and explaining failure scientifically, astrology is simply too complex for fruitful and solvable puzzles to be generated. For instance, an astrologist can make the prediction that I am going to slip on a banana peel within the next week because of my star chart. This is a risky prediction because it clearly states the conditions for verification and falsification and is potentially falsifiable. Furthermore, if the prediction fails, the astrologist may refer to other variables that exist in the complex system of astrology to explain the failure of the prediction just as a chemist could refer to other variables in a failed experiment. So, even in the framework of falsification and failure explanation, astrology can be classified as a science. 

However, Kuhn argues that, due to the sheer complexity of astrology, there is significant space for critical discourse and a lack of solvable puzzles in astrology because two astrologers looking at contradictory sets of variables may seem perfectly justified in explaining the failure of the earlier prediction only to arrive at a lack of solvable puzzles/stalemate. Where the chemist might be able to refine the process and the hypothesis-in-question to get a concrete answer as to why the experiment is behaving unexpectedly, the astrologist would not be able to do the same due to the overwhelming number of variables and sources of error that are associated with using the stars to predict someone’s future. Kuhn states that astrology is more of a craft than a science in this manner as, “theory was adequate only to establish the plausibility of the discipline and to provide a rationale for the various craft-rules” (Kuhn 1970/2012, 16). In other words, empirical data and results did not justify astrology’s existence, but astrology justified its own existence. The failures of astrological forecasts did not yield new puzzles for astrologers to solve because astrologers could often engage in critical discourse by either explaining away the failure through existing astrological frameworks or attributing the failure to one of the many sources of error. In this case, the allowance of critical discourse directly inhibits the puzzle-solving enterprise and prohibits astrology from becoming scientific. 

While both Popper and Kuhn are convinced of the pseudoscientific nature of astrology, they arrive upon this conclusion in different manners. Popper emphasizes a lack of falsifiability in astrology, whereas Kuhn uses the problem-solving nature of science to classify astrology as a pseudoscience. Furthermore, Kuhn takes the history of astrology into account to demarcate science from pseudoscience: “The history of astrology during the centuries when it was intellectually reputable records many predictions that categorically failed” (Kuhn 1970/2012, 15). Throughout history, astrologers have made many risky predictions using available theory and this fits Popper’s demarcation criterion for science. Kuhn avoids this pothole by arguing that the sheer complexity of contemporary astrology allows for critical discourse and prevents solvable puzzles from being formed within the paradigm. Thus, Kuhn has a stronger argument under the broader lens of the history of science. Effectively, Kuhn looks at science through a telescope, and Popper views science through a microscope.

Kuhn’s focus on the history of science is also apparent through his distinction between extraordinary research and normal research. By fitting Popper’s demarcation of science into the definition of the intermittent extraordinary science, Kuhn further fortifies his argument by forcing the audience to take a weighted average in favor of Kuhn’s argument for the more representative normal science. However, this raises the question of whether science should be defined by its history or its most revolutionary parts. Interestingly, Kuhn states that he would not be surprised if Popper, mistook what Kuhn calls normal science, “for an intrinsically uninteresting enterprise” (Kuhn 1970/2012, 13). However, the answer to this question is highly contextual. An aeronautical engineer looking to optimize the aerodynamics of a glider may be perfectly content solving their puzzle using traditional Newtonian dynamics. On the other hand, the traditional Newtonian conception of gravity will likely be inadequate with a modified gravity theorist attempting to explain gravity under a galactic context. In this vein, Kuhn’s and Popper’s arguments can be interpreted as two sides of the same coin. While both demarcation criteria often arrive at similar outcomes, Kuhn builds off a puzzle-solving tradition rooted in the history of science and Popper builds his demarcation criteria through testing and falsification. Though both these arguments are often seen as two sides of the same coin, Kuhn builds a more compelling argument for demarcation under the framework of the more representative normal science by taking the history of science into account, while Popper’s demarcation criteria are seemingly limited to Kuhn’s periods of extraordinary science. Thus, as Kuhn’s argument is more versatile, it paints a more compelling picture of demarcation between science and pseudoscience.

Despite these differences between Popper and Kuhn, they would probably agree on one thing: be open to the idea that your beliefs are not entirely true. Popper’s focus on falsification and Kuhn’s focus on puzzle-solving both rely on the open-mindedness of the scientist to build the realm of scientific knowledge. Throughout this essay, I use astrology as a punching bag to answer the question of demarcation of science from pseudoscience and this might be revealing of my own biases regarding the field of astrology as it exists today. However, what if future technology or realizations allow us to see through the sheer complexity of astrology and the scientific study of astrology greatly builds upon our knowledge? I would be forced to introspect as to where my beliefs have led me astray. This perspective also seems to imply that there is no universal scientific truth that can be formulated by us because we are constantly stuck in the process of generating theories that are not able to capture our observations and experiences fully. Thus, with Popper’s and Kuhn’s emphasis on open-mindedness, science (and perhaps life itself) seems more concerned with the process than it is about the destination.

Work Cited

Cover, J.A., Martin Curd, Curd, Chris Pincock (editors). 2012. Philosophy of Science: The Central Issues (Second edition). New York: W.W. Norton and Company. 

O'Connor, J J, and E F Robertson. “Mathematical Discovery of Planets.” Maths History, University of St. Andrews, Sept. 1996, https://mathshistory.st-andrews.ac.uk/HistTopics/Neptune_and_Pluto/. 

Phelan, Hayley. “Will Coronavirus Kill Astrology?” The New York Times, The New York Times, 9 May 2020, www.nytimes.com/2020/05/09/style/coronavirus-astrology-predictions.html.