The grievance studies affair

The factual record is uncontested because the authors documented it: twenty submissions, seven acceptances, four published before exposure. The dog-park paper appeared in Gender, Place & Culture, which issued an expression of concern and then withdrew it. The fieldwork described in it never happened.

What the exercise establishes is the contested part, and the criticism is methodological rather than defensive. Daniel Engber and others pointed out at the time that there was no control arm: no matched set of deliberately absurd papers submitted to journals in other fields, so the result cannot distinguish a problem specific to these disciplines from the general observation that peer review does not detect fabricated data. That general observation is well supported elsewhere — reviewers assume good faith about data they cannot see, which is why fabrication scandals occur across medicine, psychology and economics.

Mikko Lagerspetz's 2021 reconstruction in Science, Technology, & Human Values went through the submission record and found the authors' own account of it selective: several 'acceptances' were by journals of low standing, some papers were rejected repeatedly before placement, and the revision histories show reviewers demanding changes rather than waving the work through.

The project also had a real cost to its authors. Boghossian, the only one holding a university post, was found by Portland State to have conducted research on human subjects — the journal editors and reviewers — without ethics approval, which on the plain terms of the rules he had.