The traditional view, dominated by mythical revelations and uncritical dogma, started to show cracks when the Ancient Greeks introduced and developed a new method based on critical, objective observation and rational argument. The Age of Enlightenment in 17th century Europe turned these cracks into a full-blown scientific paradigm shift, founded on the Greek method and on a mechanistic, secular world view, and laid the foundations for the scientific method of contemporary science.
The scientific method relies entirely and solely on inductive empirical observation to make and justify claims about the world. The observed facts must speak for themselves, and must always have the final say. A theory about the world is sound or indeed scientific, when objectively observed instances confirm the theory. The more confirmations, the stronger the theory. Theories that cannot be consistently and coherently verified in this manner – like the mythical revelations and dogmas of old – must be rejected as non-science or pseudo-science.
Popper observes how theories like astrology, Marxism and Freudian psychoanalysis rely on this very method to claim scientific authority. But, regardless of how meaningful, important, suggestive or indeed truthful these theories may be, they cannot claim to be based on science – not in the same way Einstein’s or Newton’s theories can. Where then should the line be drawn, especially seeing that so-called scientific theories are proven wrong from time to time, while so-called non-scientific theories are verified in great numbers of observed instances? Popper suggests that if non-science can be verified by appealing to the scientific method, then surely there must be something wrong with the method. He calls this question of when a theory can be rightfully ranked as scientific, the “problem of demarcation.”
Science is an evolutionary process, a process of problem-solving and trial and error – always incomplete and inconclusive, always relying on incomplete information and knowledge, always open to refutation, or falsification. Knowing and acknowledging this, critical analysis of old theories should be central in the ongoing, evolving scientific process. It then follows that the strength of a scientific theory should not be measured in terms of how many verified phenomena we may count, but in terms of how many attempts to refute, or falsify, it is able to survive.
This acknowledgement of the fundamental uncertainties in our scientific claims about the world, was also recognised by David Hume in his formulation of the problem of induction. Hume claimed that regardless of how many observations successfully confirm a theory, our observations are always limited, and therefore unable to justify universal claims about the world. Seeing the sun set for one thousand and one times, does not justify us to claim universal certainty about whether the sun will set the next day. But should this leave us with a sceptic conclusion: that certain knowledge about the world is impossible? Popper says it shouldn’t, because while the verification test may be inconclusive, the falsifiability test allows us on the one hand to generate negative certainty by successfully refuting a theory, or on the other hand to apply increasing levels of strength to our theories, and produce the best available theory.
The problem with theories like astrology, Marxism and psychoanalysis is not that they cannot be empirically verified (they can, and they have been to some extent), but that they cannot be empirically refuted. They are so all-encompassing that any observed phenomenon can be interpreted in terms of the theory. The facts don’t speak for themselves, and don’t have the final say. Rather than that the observed phenomenon says something about the theory, the theory says something about the phenomenon – about how the phenomenon should be interpreted. Rather than confirmations of the theory, observed verifications are interpretations of phenomena in terms of the theory.
We don’t observe the world from a neutral, theory-free vantage point, but always from the perspective of particular preconceived concepts. Verifications of these concepts may be found everywhere, like self-fulfilling prophecies, if that is what we are looking for. We may not even be able to observe refuting instances, because we are looking through the lens of a particular, preconceived theory. Science, however, is looking for factual, objective and universal explanations about the world, and should always critically test our theories against the bare facts. So rather than looking for verifications that confirm our theoretical expectations, we must critically scrutinise these expectations. The best way to do this, is to submit them to attempts to refute, or falsify them. Because every scientific theory prohibits certain things from happening, it is these very things we must test in circumstances that encourage them to happen.
The initial attraction of the theories of Freud and Marx lies in their explanatory power. Each and every observed instance serves as a confirmation of the theory. But to Popper this seems to be very similar to the mythical explanations the scientific method claims to reject – in which every revelation served as a confirmation of the myth. Re-explaining a theory, or re-interpreting a phenomenon, to be able to justify a falsification makes the theory weaker, not stronger. Rather than a strength, this broad explanatory power is a weakness of a theory.
Einstein’s theories, however, are fundamentally different, in that they make risky, unexpected, predictions about the world, which can be critically, empirically tested. One unpredicted observation can and should invalidate the entire theory. This is different from Marx’ and Freud’s theories, in that their’s would be able to absorb contradictory observations by either adjusting the theory, or by explaining the observations in terms compatible with the theory. They may be right, but it is impossible to prove them wrong.