> ... > and all of this is before you get into the issues with the non trivial number of junk papers published in social science and psychology journals and the like that at best seem to say that even in cases of more difficult questions where more rigorous methods are valuable, this system of knowledge production fails a non trivial amount of the time.

As an aside, this hasn't been limited to just soft sciences like the humanities, social sciences, psychology, and the like. Going back several years, a few MIT students created a random paper generator and used it to create and submit papers that were accepted into computer science related conferences [1].

Some, many seemingly motivated by politics, have pulled similar stunts with fields related to social sciences [2]. But this ignores that the same issue is also prevalent in other areas that could be labeled as hard sciences [1]. (Of course, this is ignoring the humongous amount of papers that represent extremely niche incremental advancements that advance the state of the art in very specific but extremely limited areas.)

[1] https://news.mit.edu/2015/how-three-mit-students-fooled-scie...

[2] https://en.wikipedia.org/wiki/Grievance_studies_affair