Dan Herbatschek | Philosophy of Science

Whenever we ask a big question, it quickly breaks down into smaller ones. What makes a science a science? is no exception. To answer it, we confront the problem of demarcation: how to distinguish, in a principled and non-arbitrary way, between genuine sciences and pseudosciences.

Not every non-science is a pseudoscience. Mathematics, engineering, or the arts are serious intellectual undertakings, but no one is likely to mistake them for sciences in the same way one might confuse astrology or homeopathy with scientific inquiry. The real issue arises with fields that lay claim to science’s special authority, without actually deserving it.

Calling something a pseudoscience is not to dismiss it as wholly false or useless. Astrology, for instance, may occasionally generate claims that turn out true. The problem is that these claims are not justified in the way scientific claims are. Conversely, to call something scientific is not to guarantee its truth—scientific history is filled with bold theories later proved wrong. What “scientific” signals is something like a qualification: entry into a kind of intellectual Olympics. A scientific theory may not be the champion, but it at least deserves a place on the field. Pseudoscientific theories, on this view, should never make it to the starting line.

This matters because what counts as science has real consequences. A discipline’s claim to scientific status shapes curricula (as in battles over creationism), influences funding (think of debates over alternative medicine), and even directs social practices (such as controversies over IQ testing). Science earns a special kind of cultural authority, and we want to know what justifies granting it.

Defining Science—or Trying To

One way to secure that authority is through a definition. Philosophers mean something specific by the word: a set of individually necessary and jointly sufficient conditions. Consider the simplified legal definition of bourbon: the mash must contain 51–79% corn, and the whiskey must age at least two years in new, charred oak barrels. Each condition is necessary; together they are sufficient. This is not just cataloging common usage (the lexicographer’s job), but aiming at the essence of what makes bourbon bourbon.

A definition like this, applied to science, would make explicit the criteria we are using to sort genuine inquiry from impostors. But such definitions are notoriously hard to come by. Agreement on cases—“this counts as science, that doesn’t”—doesn’t tell us why, nor whether our reasons are good ones.

And sometimes we manage without definitions. Justice Potter Stewart’s famous remark about pornography—“I know it when I see it”—captures the idea. Adequate for some contexts, but limited. People may agree on examples but disagree on borderline cases, or even agree for very different reasons. Unsurprisingly, science and pseudoscience provoke just these kinds of disputes.

Take the Parapsychological Association’s admission to the American Association for the Advancement of Science in 1969. Was that a scientific endorsement of ESP, or a philosophical and political gesture? Without clear criteria, we cannot tell. Still, even without full definitions, we can work with partial tools. If something fails a necessary condition, we can rule it out. If it meets a set of sufficient conditions, we can rule it in. You can, for instance, know that being born in the U.S. suffices for citizenship without knowing all the possible ways of becoming a citizen.

Popper’s Radical Move: Falsifiability

No philosopher looms larger in this discussion than Karl Popper. He came of age in Vienna after World War I, immersed in a culture overflowing with artistic creativity and scientific speculation. Popper was fascinated by Einstein’s relativity, Marx’s theory of history, and the psychologies of Freud and Alfred Adler (with whom he briefly worked).

At the time, these theories of history and psychology were often presented as scientific. Engels described Marxism as an extension of Darwin into the social world. Freud likewise likened psychoanalysis to Darwin’s revolution, shifting science from biology to psychology. Popper initially took these claims seriously. But over time, he became convinced that Einstein’s relativity was different: it embraced genuine risk, exposing itself to criticism and the possibility of failure.

This commitment to criticism carried into his political thought. In The Open Society and Its Enemies (1945), written in exile from Nazi Europe, Popper argued that a society is “open” when criticism is not only permitted but effective: rulers must respond to it. The same principle defined science for him: openness to being proved wrong. (Ironically, Popper himself was famously defensive about criticism—a reminder that one can devote life to an ideal without always embodying it.)

Why not simply say that science is defined by empirical support? Because, Popper argued, observation is cheap. Every pseudoscience brims with anecdotes, case studies, or selective data. Astrology has reams of biographical and astronomical detail; Freud and Adler amassed patient histories. But none of this counts as serious testing. Observation, Popper insisted, is theory-laden: what you “see” depends on what you expect. And when theories can reinterpret every failure as a hidden success, they never risk refutation.

Thus his proposal: the mark of science is falsifiability. A scientific theory must stick its neck out—it must make bold predictions that could prove it wrong. Relativity offered a perfect case: the prediction that starlight bends around the sun’s gravity. Eddington’s 1919 eclipse expedition put this to the test. The prediction held, and relativity triumphed—but for Popper, the triumph lay less in being right than in being testable. Genuine sciences can lose; pseudosciences cannot.

The Texture of Falsifiability

Popper’s principle, though elegant, raises complexities.

• Degrees of science. He sometimes described theories as more or less scientific depending on their vulnerability to refutation. Marxism, for instance, began as a bold, predictive theory but was later shielded from counterexamples, becoming less scientific. This blurs the sharp boundary between “science” and “non-science.”
  
• Descriptive and normative. Popper claimed both that this is how scientists in fact work (at their best) and that it is how they should work.
  
• Necessary but not sufficient. Not all falsifiable claims are scientific. “Elvis will strike me down when I finish this sentence” is falsifiable (and false) but not science. Falsifiability is thus a necessary condition, not the whole story.
  
• Metaphysical value. To be unscientific is not to be worthless. Atomism, for centuries, was untestable but fruitful. Even Darwin’s principle of natural selection, Popper once argued, was dangerously close to tautology—fitness defined by survival—though later he reconsidered, seeing it as a set of historical hypotheses testable against phenomena like genetic drift.
  

Challenges to Popper

Popper’s view also faces serious objections:

1. Existential claims. “There exists a gold sphere a mile wide somewhere in the universe” cannot be falsified by finite search but does not seem unscientific. Popper’s reply: science cares about laws (e.g., “All copper conducts electricity”), which can be falsified by a single counterexample.
   
2. Probabilistic statements. Much science is statistical. Fifty sixes in a row does not falsify the claim that a die is fair. Evolutionary biology relies heavily on probability, too. Popper suggested scientists could adopt agreed thresholds for when to treat a claim as falsified, but that reduces the criterion to convention.
   
3. Theory versus practice. Is scientific status a property of the theory itself or of how it is handled by practitioners? One Marxist might cling to failed predictions; another might revise or abandon them. Popper often spoke as if falsifiability were a logical property of the theory, but practice complicates this.
   
4. The tenacity of good theories. Scientists do not discard good theories at the first anomaly. Astronomers did not abandon Newtonian physics when Uranus’s orbit deviated; they posited Neptune, later confirmed. Some auxiliary moves are legitimate; others are ad hoc. Popper must explain the difference.
   
5. Explanatory scope. Many philosophers see wide explanatory power and empirical confirmation as genuine virtues. Popper’s single-minded focus on risk-taking sometimes seems to undervalue these.
   

Where We End Up

Popper gave us a powerful insight: science distinguishes itself by courting refutation. Bold conjectures, exposed to failure, separate genuine inquiry from pseudoscience. Yet falsifiability alone is not enough to solve the demarcation problem. Science also relies on explanatory scope, predictive accuracy, reproducibility, independent testability of auxiliaries, and responsiveness to criticism.

Still, Popper’s criterion remains one of the most elegant and influential answers we have. It reframes science not as the pursuit of confirmation but as a willingness to risk being wrong—and that, perhaps more than anything, captures the spirit of the enterprise.