We’re living in a time of falsehood. Between ‘alternative facts’ and a disturbing preponderance of fake news, the lack of common ground truth precludes productive conversation. I don’t have the solution to this. But I believe a small part of the solution is understanding the logical framework for separating science from untestable, irrefutable claims. To that end, dear reader, I’d like to tell you about the philosophy of science.
If your secondary education was anything like mine, then you were probably (implicitly) lied to about science. For me, science was presented as a ‘body of truth’ — a set of equations and facts to memorize and regurgitate for the exam. In college this view got a bit more nuanced — the equations we worked with were good approximations in particular settings, but lacking compared to some underlying truth. My mental model of science was still one of verification though — propose a hypothesis, then confirm it through experiment or rework if necessary. This model, however, is totally wrong.
This Verificationist perspective on science trips on a very old philosophical issue: the Problem of Induction. To illustrate — Europeans used to believe that all swans are white. They had observed many swans, found only white samples, and drew a generalized conclusion that all swans are white. But when they eventually explored Australia, they made a discovery which caused their theory to crumble: black swans. The problem is in thinking a theory confirmed by limited observations; a ‘proven’ theory purports to hold for an infinite set of cases, but experiments are necessarily limited to a finite number. How can we logically place trust in scientific theory?
Enter Karl Popper with his 20th century work “The Logic of Scientific Discovery”, where he laid out (among other things) the idea of Falsification. Popper argued that science should not be conducted by attempting to prove theories true, but rather by attempting to show them to be false — to falsify them. In this approach, a theory is never accepted as truth, but instead gains more credibility as it fails to be shown false.
Practically speaking, why does this matter? First, it gives a logically sound formulation for doing science. Second, it allows us to separate science from ‘pseudo-science’. Popper was a contemporary of Freud, and noted that Freudian Psychology was flexible enough to incorporate any new observations; Freud claimed to be able to explain any behavior of a female in terms of penis envy. This sort of theory is not falsifiable, and therefore does not deserve to be called science. A scientific theory disallows particular behaviors (e.g. perpetual motion); such a theory has some predictive capability.
Admittedly, this doesn’t solve the problem of outright falsehood. If we don’t agree on what observations were made, then the inferences we draw will be completely different. But understanding Falsification allows us to determine what kind of claims can be refuted, and separate out junk purporting to be science. If I’ve piqued your interest in the subject, I recommend both Popper’s book (a dense read) and the Crash Course YouTube series on Philosophy (an absolute joy).