The coronavirus variolation theory hinges on two assumptions that are difficult to prove: that lower doses of the virus lead to less severe disease, and that mild or asymptomatic infections can spur long-term protection against subsequent bouts of sickness. Although other pathogens offer some precedent for both concepts, the evidence for the coronavirus remains sparse, in part because scientists have only had the opportunity to study the virus for a few months.

Experiments in hamsters have hinted at a connection between dose and disease. Earlier this year, a team of researchers in China found that hamsters housed behind a barrier made of surgical masks were less likely to get infected by the coronavirus. And those who did contract the virus became less sick than other animals without masks to protect them.

A few observations in humans seem to support this trend as well. In crowded settings where masks are in widespread use, infection rates seem to plummet. And although face coverings cannot block all inbound virus particles for all people, they do seem to be linked to less illness. Researchers have uncovered largely silent, symptomless outbreaks in venues from cruise ships to food processing plants, all full of mostly masked people.