Questions: Falsifiability as the Criterion of Demarcation

Question 1

A new theory of personality psychology has been confirmed by thousands of case studies, but it is formulated so that it can explain any personality trait or behavioral pattern after the fact. Under Popper's demarcation criterion, how should this theory be classified?

• A. Scientific, because it has been confirmed by a large and diverse body of evidence
• B. Scientific, because it makes accurate predictions about human behavior
• C. Non-scientific or pseudoscientific, because it cannot specify which observations would refute it — a theory that explains everything predicts nothing
• D. Non-scientific only if it has been explicitly used to make false predictions

Answer: (C)

Popper's criterion does not reward confirmation — it demands falsifiability. A theory that can accommodate any possible observation makes no genuine claim about what we will observe. Popper pointed to Freudian psychoanalysis as an example: whatever a patient did, the analyst could explain it as confirming the theory. The number of confirmations is irrelevant; what matters is whether the theory could in principle be refuted by an observation it did not predict.

Question 2

When early astronomers found that Newtonian mechanics failed to fully account for Uranus's orbit, they postulated a new planet (eventually confirmed as Neptune) rather than abandoning Newton's laws. What does this illustrate about falsification in practice?

• A. It shows that Newtonian mechanics was not falsifiable and therefore not scientific
• B. It illustrates the Duhem-Quine problem: when a prediction fails, you can save the core theory by modifying an auxiliary hypothesis rather than rejecting the theory itself
• C. It proves that scientific theories are never actually falsified — scientists always preserve them
• D. It shows that confirmation (discovering Neptune) is what really drives science, not falsification

Answer: (B)

The Duhem-Quine problem shows that no single theory is tested in isolation. When a prediction fails, the blame could fall on any auxiliary assumption rather than the core theory. Saving Newton's laws by adjusting the auxiliary (that all relevant bodies had been accounted for) was scientifically productive here because it led to a novel prediction — Neptune's existence — that was confirmed. Lakatos distinguished progressive research programmes (generating new successful predictions) from degenerative ones (only accommodating old failures) to capture why this move was legitimate in Newton's case.

Question 3

According to Popper, a theory that makes specific, risky predictions — ones that could easily be shown false — is more scientific than a theory that only predicts what is already likely.

• T. True
• F. False

Answer: True

This is the core of Popper's asymmetry between confirmation and falsification. A 'risky' prediction — unlikely if the theory were false — is precisely what distinguishes bold scientific theories from vague ones. Einstein's prediction that light bends around the sun by a specific, quantified amount was risky: it was an unusual claim that could easily have been refuted by measurement. Theories that only predict what is likely regardless of their truth have no real empirical content. The more a theory rules out, the more scientific it is.

Question 4

The Duhem-Quine problem shows that any theory can be protected from refutation by adjusting auxiliary hypotheses, proving that falsifiability cannot usefully distinguish science from non-science.

• T. True
• F. False

Answer: False

The Duhem-Quine problem complicates falsifiability as a sharp criterion, but it does not eliminate its usefulness. Lakatos's response — distinguishing progressive research programmes (which use auxiliary adjustments to generate novel successful predictions) from degenerative ones (which only patch up past failures) — preserves the spirit of Popper's demarcation. Falsifiability becomes less a binary criterion and more a measure of methodological integrity: are auxiliary adjustments made in advance with testable consequences, or only post-hoc to avoid refutation?

Question 5

Why does Popper's criterion classify Freudian psychoanalysis as pseudoscience, even though it provides rich explanations for a wide range of human behaviors?

Model answer: Popper's objection is that psychoanalytic theory is formulated so it can explain any behavior after the fact — but this explanatory power is precisely what disqualifies it. If a patient is aggressive, the analyst explains it as expression of repressed impulses; if passive, as suppression of those same impulses. No patient behavior could count as evidence against the theory. A theory compatible with every possible observation has no empirical content — it tells us nothing about what we will actually observe. Popper contrasted this with Einstein's general relativity, which made specific, risky predictions that could have been refuted.

Any sufficiently flexible framework can explain past events. Scientific theories earn their status by making specific claims that constrain what we will observe, thereby risking refutation. Psychoanalysis's ability to explain everything is its weakness, not its strength. The key asymmetry: a million confirmations can't prove a universal claim, but one genuine counterexample disproves it — so the scientific attitude is 'what would prove this wrong?' not 'how can I confirm this?'