A traditional view is that scientific evidence can be produced only by
intersubjective methods that can be used by different investigators a
nd will produce agreement. This intersubjectivity, or publicity, const
raint ostensibly excludes introspection. But contemporary cognitive sc
ientists regularly rely on their subjects' introspective reports in ma
ny areas, especially in the study of consciousness. So there is a tens
ion between actual scientific practice and the publicity requirement.
Which should give way? This paper argues against the publicity require
ment and against a fallback version of it, viz. that evidence-conferri
ng methods must at least have their reliability publicly validated.