I think roughly conscientious scientists already try doing that today, that’s why I’m saying this isn’t a hypothetical. Having a stronger norm might lead to more of them doing that; some defectors would always remain of course, but maybe they would at least be regarded less well within their own community. Forget journalists, right now we have a problem even with academic journals being biased towards catchy positive results.
if it’s fairly common for even good scientists to give their papers such titles
Depends also about what we’re talking about. Honestly my experience is that the most stand out type of paper title isn’t necessarily “sensational claim” as much as “tongue in cheek reference”, which is a pretty neutral concession to visibility. It’s a very field-dependent problem, though. But title-wise, I think the worst of the drift happens in press releases and then through journalists. You can make paper titles catchy without making them state any claim (easy pattern: just state WHAT you’re looking at, like “Studying the relationship of X and Y” instead of “Positive correlation of Y with X” or whatever).
I think roughly conscientious scientists already try doing that today, that’s why I’m saying this isn’t a hypothetical. Having a stronger norm might lead to more of them doing that; some defectors would always remain of course, but maybe they would at least be regarded less well within their own community. Forget journalists, right now we have a problem even with academic journals being biased towards catchy positive results.
Depends also about what we’re talking about. Honestly my experience is that the most stand out type of paper title isn’t necessarily “sensational claim” as much as “tongue in cheek reference”, which is a pretty neutral concession to visibility. It’s a very field-dependent problem, though. But title-wise, I think the worst of the drift happens in press releases and then through journalists. You can make paper titles catchy without making them state any claim (easy pattern: just state WHAT you’re looking at, like “Studying the relationship of X and Y” instead of “Positive correlation of Y with X” or whatever).