Finnur Dellsén: "Collectivizing Scientific Testimony".

On many scientific topics of public interest, there are multiple scientific experts who are more-or-less equally reliable in evaluating whether a given scientific claim is correct. This talk addresses two closely related questions about such situations: First, how should laypeople form or update their opinions on such topics in light of the experts’ opinions? Second, what should the scientific experts report to laypeople? Regarding the first question, I argue that laypeople should generally not form or update their opinions by deferring to the positions taken by individual scientific experts; rather, they should consider the expert opinion distribution more generally and use various heuristics for forming reliable opinions on that basis. Regarding the second question, I argue that scientific experts should not report their own position on the topic; rather, they should provide information about the expert opinion distribution more generally, e.g. by communicating that a majority of their fellow experts take such-and-such position on the topic, even when they personally disagree with that position. I then go on to trace some practical implications of these arguments, e.g. that scientific communities have an obligation to periodically gather data on the expert opinion distributions regarding topics that are of interest to the general public.