The problem of causality has been a contentious topic of discussion in the philosophy of science. 18th-century philosopher David Hume thought that our idea of causality came from seeing “constant conjunctions” of events happening in sequence. By observing the same sequence of events repetitively we begin to think that the latter event necessarily follows from the former, and that in the absence of the preceding event, the second event will not follow. However, no finite number of observations can guarantee the outcome of all possible cases. This form of reasoning — using past observations to make general, future predictions — is called induction, and Hume believed it to have no logical basis despite being central in the sciences.
Two centuries later, Austrian-British philosopher Karl Popper questioned Hume’s conception of causation. Hume’s argument that we get causality from constant conjunctions is insufficient to account for the conscious expectation that we feel when we think we are causing something, since repetitive observation of conjugated events would make the process of identifying it physiological and unconscious. Popper believed that science does not operate on inductive reasoning from constant conjunctions, but from conjectures that are tested by repeated observations and experiments. These conjectures do not derive from inductive reasoning but from a process “of actively trying to impose regularities upon the world.” Science doesn’t proceed from observation to theory; it begins with theory — conditioned by our genetic and psychological preferences — that is used to discern observations that count as meaningful. From these selected observations, a new theory is born.
These investigations on the idea of causation remind us that science is not an exceptionless field that supposedly “does not care about your feelings.” In fact, what we know as science today is the result of preferential observations coming from the tendency to adhere to the initial theory. Furthermore, we proceed from observation to new theory not on the basis of a logical connection between the two, but by “jumping” first to a theory, and then repeatedly testing it with critical methods. Because the generalization from observation to theory is unwarranted, the latter must be investigated under risky predictions that expect an event to refute the theory. The “confirmation” of Einstein’s theory of general relativity by the 1919 Eddington Experiment was such a case. The theory predicted the sun to have an angular deflection — the bending of starlight around the sun — of 1.75 arcseconds during a solar eclipse, which was twice the value predicted by the Newtonian theory of gravity. The Newtonian theory having been the predominant one at that time, an observation consistent with the Einsteinian prediction was an abnormality that couldn’t have been explained without in light of a drastically new theory; it was a risky prediction and therefore a meaningful one. While the riskiness of a prediction does not demarcate science from non-science, it sheds light to the fact that not all science is created equally.
~ Benjamin Ha `27
Be First to Comment