> If evidence is produced that goes against a consensus then the consensus is wrong To be more specific, then the consensus may be wrong. The bar to demonstra

> If evidence is produced that goes against a consensus then the consensus is wrong To be more specific, then the consensus may be wrong. The bar to demonstrate that something is wrong and be taken seriously is meant to be higher than simply claiming one can produce evidence to the contrary. When such evidence is discovered, the idea is that one would present that evidence to peers, who would attempt to replicate the findings, and support them if they are able to do so, or otherwise publish their counter-findings in response. This is obviously more complicated in practice because we're dealing with humans and not ideals (and thus we have things like the replication crisis), but that is the theory of how we advance the body of knowledge that we derive from scientific theses and experiments. We do not advance it by claiming that evidence has been found and then leaving it at that, in a vacuum. So it's not exactly that science "works on consensus", in the sense that just a majority doesn't prove anything, but -- convincing, using evidence, the majority of the scientific community in your field, is certainly part of the process.