Of course we'd like researchers to always start from the best premises. But the process is such that if you want to update your premises, you have twenty compet

Of course we'd like researchers to always start from the best premises. But the process is such that if you want to update your premises, you have twenty competing theories to choose from, all in the process of getting validated. Do you move and where do you move? When I read this article I immediately felt with the author how these people should update their models. How long can it take? But then again, how do I know the author's choice studies are the good ones? If I were to talks to somebody in the field, they'd note that those studies are "promising" or "concerning". But maybe they won't change what they're doing just yet. Inertia? Yes. Bad? Oh, where's hindsight when you need it the most! The book The Golem by Collins and Pinch discusses historical examples of the scientific process for things we take for granted today. They show how shaky the confirming experiments were. How they could've easily gone the other way, and how some positive early results were based on selecting favourable data.