Karl Popper, in his paper "Science: Conjectures and Refutations", seems to be making accusations more than he is describing aspects of the scientific method. In this paper, written in 1953 at the height of McCarthyism, he dismisses Marxism by associating it with psychologists he doesn't like. There are legitimate ways to argue against Marxism, but guilt by association is a not a useful economic discussion.
He's on more solid ground when he elaborates on his theory of falsifiability. "A theory which is not refutable by any conceivable event is nonscientific." That only works if science is a process (as opposed to an event) since there are sciences that rely on studies of events that cannot be replicated, for example archeology, and for which the only test is further research. Still, I can accept that a scientific theory must be testable by trying to prove it wrong. Merely offering corroborative evidence in support does not make a theory science. (Alternately, shooting down parts of other theories doesn't mean yours is necessarily science either; the Creationists fall into this trap, but that discussion is beyond the scope of this paper.)
More problematic is "Every genuine test of a theory is an attempt to falsify it, or to refute it." Popper assumes that every test has a binary outcome: Yes or no, up or down. Certainly, any set of tests that doesn't allow for a negative result is a bad set of tests. And yet, not all questions are Yes Or No questions. Some questions require specificity, not falsifiability.
"What channel is Bugs Bunny on?" has an answer that is not yes or no. It's possible that Bugs isn't on, and scanning the channel listings may produce a negative result. But that wasn't your question; you assumed a different (though related) question was answered in the affirmative, "Is Bugs Bunny on tv?"
Let us try a more scientific question: "What causes earthquakes?" We know something causes earthquakes. Our theory works if it explains (and predicts) quakes. A theory on earthquakes may be refuted, but it is not inherently a yes or no question.
Let us create another experiment. We wish to determine if Newton's Theory of Gravitation is correct, and that a rock will fall at the speed predicted by this theory, with an acceleration of 32 ft. per second per second.
We start by holding a rock in one hand and a stopwatch in another. We let go of the rock. It falls and hits the ground at the predicted time. We test this at different heights and locales, and the theory works.
Once, we let go of the rock and it doesn't move at all. Has Newton been disproved? We examine the factors involved and determine that the rock was sitting on a table when we let go. We know something more, and have a more specific answer: A rock will fall at the speed Newton predicted when it is allowed to fall.
We try again, and find the rock suddenly shoots up and away. Has Newton been disproved? We examine the factors involved and determine that a tornado has influenced the rock. We know something more (much to the relief of future research assistants) and have a more specific answer: A rock will fall at the speed Newton predicted when it is allowed to fall with nothing else acting on it. In addition to Newton's Law of Gravitation, we have demonstrated Newton's First Law of Motion.
We try again, and the rock doesn't move at all. There's no table and no wind. Has Newton been disproved? We examine the factors involved and determine that the rock is in outer space, and there's not enough gravity for the rock to be affected. Newton (along with Kepler) was right about that too; distance between gravitational objects is a factor. In fact, the rock is falling, but it has a lateral motion too. It's in orbit. We were not at rest, with respect to the center of gravity; we were circling the planet. Newton's laws predict the necessary speed of orbits at various distances from the Earth, and this experiment has become a data point for that set of theorems in addition to ours. We know something more and have a more specific answer: A rock will fall at the speed Newton predicted when it's when it starts off at rest and is allowed to fall with nothing else acting on it.
We could (and should) test these laws further. But so far at no point have we run into Popper's criterion of refutability. We might be wrong, but that lurks in the background of our experiments. Refutability is important to testing, but it is not the only aspect of science worth testing.
Instead of falsifiability, I propose specificity: Using the built in feedback loop in the scientific method to continue to hone our experiments. Asking more specific questions to develop more specific hypotheses that lead to tests that can then be used to ask more specific questions. Being more specific helps isolate the precise cause of an effect. Along the way, we may find that our original hypothesis needed to be thrown out, but we may find that our hypothesis needed only to be altered.
Ed. Note: The instructor's comment was "an interesting and (as far as I know) a novel theory of science. Intriguing." I recieved a 49 out of 50 for the paper.
Back to main Rhetoric Page