EXPERIMENTS · MACHINE LEARNING · STATISTICS
EXPERIMENTS · MACHINE LEARNING · STATISTICS
A key feature of Merton's 1942 account of science is disinterestedness, which seems analogous to something like being willing to evaluate new claims from behind a veil of ignorance, setting aside sectarian divides over “school of thought”. Of course science has a particular status in the outside world, which can put disinterestedness at risk. In thinking through this dynamic Merton explores a set of issues I think 1) is important and 2) remains under-discussed in the literature. This brings us to the notion of scientific truth. Merton connects the status of science to the technological advancements science has facilitated; the changes in material conditions science allows for. I think this is the right move, rather than focusing abstractly on “knowledge”. To quote the mathematician Vladimir Voevodsky: “first order arithmetic may be inconsistent but the planes will still fly.” It may be useful to go further. Perhaps we should take some rough constructivist metaphysics (there is the world we carry around in our heads, which constitutes the world we experience, and there is a separate external material world, but we cannot know anything about that external world) and then move on. Then we might replace a notion of science’s value as some vague and un-analyzable epistemic guarantee with a careful analysis of science’s impacts on people’s lives and people’s various (collective) capacities.
In the same breath Merton also discusses how people often attempt to co-opt this authority to make various claims. We might call this the political economy of scientistic (as opposed to scientific) rhetoric. Merton highlights how scientistic myths, particularly on race and the economy, can resonate strongly precisely because they can be untethered from constraints of careful measurement and skeptical interrogation, and so can hew closer to “common sense” (perhaps even in Gramscian terms). This seems like a really incisive point: truths that are made up and mobilized politically can afford to sound more plausible precisely because they need not be be based on empirical evidence and contorted to capture the complexity of the world. Of course politics can infuse science in degrees rather than in totality. Trump claiming that bleach cures covid is baseless, but there is “real doubt” about climate science. Or at least there was. Instead of demagogues and bully pulpits people with power can invest in constructing a base of scientific evidence, a set of scientific facts they can point to, to support their interests as. A final and even more difficult to discern way for science to really become enmeshed irrecoverably from politics is in cases of serious underdetermination. When there is simply not enough evidence, not good enough measurements, not enough data, no external check to ensure that results remain grounded, then interests of those conducting the science can shine through. In fact those with very strong interests will seek out these underdetermined corners and use them as an opportunity to stake political claims. I would argue eugenics and economics have played this role a number of times, fittingly following Merton. This set of ideas is of course interesting to bring into dialogue with Bloor, but I am out of time and out of words.