Review #34: The Logic of Scientific Discovery

Scientific proof that nothing scientific can be proved

The Logic of Scientific Discovery

Karl Popper’s 1934 book The Logic of Scientific Discovery is a complex work. When I was narrowing down books for the reading project, I had to choose between this text and Einstein’s famous writings on special relativity. Recalling my poor showing in my college physics class, I choose Popper’s book. It turns out that book complexity is relative, since this classic text is absolutely one of the most difficult and challenging things I have ever read.

The core premise of the book isn’t that bad. The author simply wants to ensure that science is done in the most scientific and accurate way possible. The key, he insists, is falsifiability. To use this method, scientific statements need to be worded as “basic, universal, non-existence statements.” “Most ravens are black” is a poor statement because it can never be proved; there is no way of knowing that a giant stash of white ravens isn’t hiding on a moon of Jupiter. Instead, testable scientific statements need to be “there-is-not” in form: “There is no perpetual motion machine” or “There is no raven that is white.” When stated this way, it is possible to prove the statement false by demonstrating one conflicting example. Show up with a perpetual motion machine or a white raven in your hand and you’ve produced a firm response to the original hypothesis.

For you logic buffs, Popper’s system is based on “denying the consequent,” the idea that if P implies Q, and Q is false, then P is also false. I never studied logic at this level, and it doesn’t fully make sense to me. But the thing about perpetual motion machines seems sound. Popper goes on to formally establish his idea using symbolic logic, mathematics, linguistics, philosophy, and advanced physics. Those of us with IQs below 180 can commence cowering in the corner.

Popper’s goal in the text is an admirable one: “to analyze the method of the empirical sciences.” That is, he wants to determine when a statement or theory is scientifically valid, and at what point a hypothesis is proved.

It turns out that the right answer, according to Popper, is: never. No theory can ever be verified, although they can be “corroborated.” The problem is that every statement must be placed at the feet of falsifiability, and we sometimes lack the skills, the resources, or the total energy output of the universe needed to test every natural law. Even if you are able to confirm some core idea, it can always be broken down into simpler, more basic ideas that in turn must be tested.

Fortunately, Popper does not require that any testing be performed. “I do not demand that every scientific statement must have in fact been tested before it is accepted. I only demand that every such statement must be capable of being tested.”

The book’s 85 chapters consume less than half the total pages. The voluminous appendices provide additional support for his system, plus a response to the book written by Einstein himself. I found it curious that Popper went to all this trouble to prove his point, since one major theme of the book is that nothing is ultimately certain. As he says in the book’s conclusion: “The old scientific ideal of episteme—of absolutely certain, demonstrable knowledge—has proved to be an idol. The demand for scientific objectivity makes it inevitable that every scientific statement must remain tentative forever. It may indeed be corroborated, but every corroboration is relative to other statements which, again, are tentative. Only in our subjective experiences of conviction, in our subjective faith, can we be ‘absolutely certain.'”

The Well-Read Man Project

For more information about this book, visit its Well-Read Man Project page.


Please enter your comment!
Please enter your name here