Anglo-American philosophy of science has been theory-centric since at least the dominance of scientific realism in the late 20^th century. While more recent work has tended away from “general” philosophy of science and toward “foundational” questions specific to particular sciences, and has focused more on experimentation and actual scientific practices, the attitude that everything in science begins and ends with theory, and its confirmation and progress, remains stubbornly recalcitrant (although now is more surreptitiously expressed).

A recent focus on experiment tools and their patterns of development in laboratory-driven sciences like neurobiology challenges this theory-centrism. Tools that revolutionized neuroscience, at least in the eyes of neuroscientists, developed by way of atheoretical tinkering in the laboratory—by solving engineering and applied science problems, by trial-and-error, and even by seer serendipity—and not by the systematic application of theory. A common general pattern runs through these cases: laboratory tinkering -> new experiment tools and designs -> theory progress. I’ll argue for this pattern by way of some historical facts about the development of a number of neurobiology’s most influential experiment tools.