Actually, when these theories are in competition researching phlogiston looks exactly like researching the new chemistry. What I mean is that even scientists holding on to the phlogiston theory will be aware of the results that favor the new chemistry and will design experiments specifically so that the results expected by one theory will be easily distinguishable from the predictions of the other theory. As evidence piles up, both theories will be modified by their adherents to explain the experimental results; the worse theory will require more modification but the better one probably wasn’t perfect. Eventually someone writes a big review paper that summarizes the work in the field and comes out strongly in favor of oxygen-based theories; if there’s no serious further debate the writers of future textbooks will refer to the big review paper.
I’m not sure whether this is true of chemistry, but the research process you describe certainly sounds plausible. As you say, there may be many cases in which the distribution of labor doesn’t matter, because researching different theories looks the same. One area in which researching different theories looks different is research into what killed the dinosaurs. Producing geological evidence relevant to the hypothesis that volcanoes killed the dinosaurs means digging at different sites from those you would investigate for evidence about whether a meteor killed the dinosaurs. There might be sites that contain evidence relevant to both, but for the most part, research is planned by using the data we already have to look at sites where we expect to find volcanic ash or meteorite craters.
If everyone was satisfied with the answer that the meteor killed the dinosaurs, we’d miss out on the ways that volcanic activity contributed to the extinction event.
Actually, when these theories are in competition researching phlogiston looks exactly like researching the new chemistry. What I mean is that even scientists holding on to the phlogiston theory will be aware of the results that favor the new chemistry and will design experiments specifically so that the results expected by one theory will be easily distinguishable from the predictions of the other theory. As evidence piles up, both theories will be modified by their adherents to explain the experimental results; the worse theory will require more modification but the better one probably wasn’t perfect. Eventually someone writes a big review paper that summarizes the work in the field and comes out strongly in favor of oxygen-based theories; if there’s no serious further debate the writers of future textbooks will refer to the big review paper.
I’m not sure whether this is true of chemistry, but the research process you describe certainly sounds plausible. As you say, there may be many cases in which the distribution of labor doesn’t matter, because researching different theories looks the same. One area in which researching different theories looks different is research into what killed the dinosaurs. Producing geological evidence relevant to the hypothesis that volcanoes killed the dinosaurs means digging at different sites from those you would investigate for evidence about whether a meteor killed the dinosaurs. There might be sites that contain evidence relevant to both, but for the most part, research is planned by using the data we already have to look at sites where we expect to find volcanic ash or meteorite craters.
If everyone was satisfied with the answer that the meteor killed the dinosaurs, we’d miss out on the ways that volcanic activity contributed to the extinction event.
Fair enough. I’m a chemist by training, so I described what I know.