Gather a crowd of historians and philosophers of science into a room and ask them to define “science.” On second thought, don’t try this at home, because you’d likely meet with stony-faced refusal on the part of the first and raucous disagreement from the second. Yet isn’t the task rather straightforward? Isn’t this just another classic instance of academics creating mountains out of molehills? Actually, no. The problem is fiendishly frustrating (and likely intractable) simply because of the kind of activity science actually turns out to be in practice.
Consider, for example, what it clearly isn’t. Science cannot be simply a collection of true propositions about nature. Most of what has counted uncontroversially as “science” during the past few centuries—geocentric astronomy, phlogiston chemistry, ether physics, the inheritance of acquired characteristics— is now considered to be false. Even worse, much of what we now consider to be science is doubtless going to be proven false, since nature was unkind enough to deny us the answer key. Science is also not merely the proper execution of method, both because various disciplines display a whole hodgepodge of different methods, and also because one can apply all the accepted methodology and come up with doctrines (parapsychology, eugenics, phrenology) that we would with alacrity exclude. The problem gets worse when you go farther back in time or across cultures. Mayan astronomy, Classical Chinese alchemy, Hippocratic medicine—all these are rather distinct from what we now consider to be “science,” and yet it strikes most scholars as rather churlish to dismiss them. No one has been able to come up with a broadly consensual definition of science, and I am certainly not about to do so here.f